Pricing Updates

This page lists, newest first, each model’s launch pricing and the price changes that followed. Launch entries show the model’s current unit price (earlier prices are in the corresponding “Price change” entries); retired models are kept and marked with their retirement date. Pricing changes apply only to calls made after the change takes effect; historical calls are unaffected. Each model’s live unit price is shown in the Models overview and the console “Model Market”; check each call’s cost fields and units against Request & Response Format and the relevant endpoint documentation.

Pricing change

seedance-2.0-mini and seedance-2.0-fast rates raisedPer-second rates by tier:

Model	480p	720p
`seedance-2.0-mini`	$0.012655 → $0.031639	$0.027216 → $0.06804
`seedance-2.0-mini` with reference video / input video per second	$0.007593 → $0.018983	$0.01633 → $0.040824
`seedance-2.0-fast`	$0.04482 → $0.050622	$0.0963 → $0.108864
`seedance-2.0-fast` with reference video / input video per second	$0.02664 → $0.029831	$0.05769 → $0.064152

The change takes effect immediately and applies only to calls made after it.

Model launch

7 models launched · pricingText

Model ID	Price (USD)
`claude-haiku-5-5`	Input $0.09 · output $0.45 · cache read $0.009 · cache write $0.1125 · cache write (1h) $0.18; above 100K input tokens: input $0.45 · output $2.25 · cache read $0.045 · cache write $0.5625 · cache write (1h) $0.9 (per 1M tokens)

Image

Model ID	Price (USD)
`flux-3-image`	Per image: 768sq $0.0369 · 1k $0.0432 · 1.5k $0.063 · 2k $0.09 · 4k $0.5463; reference images are free
`gemini-nano-banana-2.1`	Actual token usage: text input $1.35 · image input $1.35 · text output $6.75 · image output $27 (per 1M tokens)
`gemini-nano-banana-2.1-rev`	Per image: 1k $0.0225 · 2k $0.028125 · 4k $0.03375
`mai-image-2.6`	Actual token usage: text input $4.5 · image input $7.2 · image output $34.2 (per 1M tokens)
`mai-image-2.6-flash`	Actual token usage: text input $1.575 · image input $2.25 · image output $17.1 (per 1M tokens)

Video

Model ID	Price (USD)
`vidu-q4-preview`	Per second: 540p $0.0504 · 720p $0.1026 · 1080p $0.108 · 2k $0.171 · 4k $0.351, same with or without audio; reference images and audio are free

Pricing change

gpt-6-sol long-context output rate raised

Model	Change
`gpt-6-sol`	Output above 272K input tokens: $15 → $15.75 (per 1M tokens)

All other rates are unchanged. Changes take effect immediately and apply only to calls made afterwards.

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`claude-sonnet-5-5`	input $1.8 · output $9 · cache read $0.18 · cache write $2.25 · cache write (1h) $3.6 (per 1M tokens)
`gpt-6.1-sol`	input $1.8 · output $9 · cache read $0.18 · cache write $2.25; above 272K input tokens: input $3.6 / output $15.75 (per 1M tokens)

Model launch

minimax-h3-max launched · pricing

Model ID	Unit price (USD)
`minimax-h3-max`	Per second: 480p $0.045 · 768p $0.072 · 1080p $0.144, same with references; reference video per second 480p $0.04977 · 768p $0.1287 · 1080p $0.13428; first 2 reference images free, $0.0666 per image from the 3rd

Pricing change

Video: Veo / Kling 4K, Grok video editing, Wan 3.0 input-video pricing; PixVerse, Seedance, Hailuo and Vidu repricedNew tiers

Model	New unit price
`veo3.1-fast`	Per second: 4K $0.225 · 4K with audio $0.27
`veo3.1-quality`	Per second: 4K $0.36 · 4K with audio $0.54
`veo3.1-lite`	Per clip: 4K $0.64125
`veo3.1-fast-rev`	Per clip: 4K $0.72
`veo3.1-quality-rev`	Per clip: 4K $1.6875
`kling-v3`	Per second: 4K / 4K with audio $0.48213
`kling-v3-omni`	Per second: 4K / 4K with audio / 4K with reference video $0.48213
`grok-imagine-video`	`edit`: output $0.063 per second + source video $0.009 per second (same at 480p / 720p)
`wan3.0-video`	Input video per second: 480p $0.045 · 720p $0.09 · 1080p $0.18
`wan3.0-video-prime`	Input video per second: 480p $0.0612 · 720p $0.126 · 1080p $0.252

PixVerse V6 (per second, increased)

Tier	Silent	With audio
360p	$0.018 → $0.0225	$0.027 → $0.0315
540p	$0.027 → $0.0315	$0.036 → $0.0405
720p	$0.036 → $0.0405	$0.045 → $0.054
1080p	$0.072 → $0.081	$0.09 → $0.1035

Seedance (per second)

Model	480p	720p	1080p	4K
`seedance-1.0-pro`	$0.0216 → $0.02187	$0.0468 → $0.046332	$0.1044 → $0.11016	—
`seedance-1.0-pro-fast`	$0.009 → $0.0099	$0.02 → $0.0225	$0.0432 → $0.044064	—
`seedance-1.5-pro` (same with or without audio)	$0.0216 → $0.021695	$0.0468 → $0.046656	$0.116 → $0.1215	—
`seedance-2.0`	$0.07 → $0.07425	$0.15 → $0.15975	$0.37 → $0.3987	$0.78 → $0.81225
`seedance-2.0-fast`	$0.0405 → $0.04482	$0.09 → $0.0963	—	—
`seedance-2.0-mini`	$0.0144 → $0.012655	$0.0288 → $0.027216	—	—
`seedance-2.0-mini` with reference video / input video per second	$0.0072 → $0.007593	$0.01557 → $0.01633	—	—
`seedance-2.5`	$0.103 → $0.10809	$0.231 → $0.243	$0.409 → $0.511758	—
`seedance-2.5` with reference video / input video per second	$0.072 → $0.0648	—	$0.25866 → $0.30618	—

On seedance-2.5, billed input-video seconds are at least 0.75 × output seconds.Hailuo (per second → per clip)

Model	Tier	Old (per clip equivalent)	New (per clip)
`minimax-hailuo-02`	512p 5 s	$0.0585	$0.09
	512p 10 s	$0.117	$0.135
	768p 5 s	$0.225	$0.252
	768p 10 s	$0.45	$0.504
	1080p 5 s	$0.45	$0.441
`minimax-hailuo-2.3`	768p 6 s	$0.3294	$0.3294
	768p 10 s	$0.549	$0.504
	1080p 6 s	$0.486	$0.441
`minimax-hailuo-2.3-fast`	768p 6 s	$0.1674	$0.171
	768p 10 s	$0.279	$0.288
	1080p 6 s	$0.2862	$0.297

Vidu (per second, reduced)

Model	720p	1080p
`vidu-q3-mix`	$0.1125 → $0.108	$0.135 → $0.1305
`vidu-q3-turbo`	$0.054 → $0.0495	$0.063 → $0.0585

Wan

Model	Change
`wan2.7-r2v`	Reference video billed for at most 5 seconds
`wan2.7-videoedit`	6-second minimum on output removed

seedance-2.5, wan3.0-video, wan3.0-video-prime and grok-imagine-video freeze quota from the request at submission and settle on the actual result when the task completes; any excess is refunded.Changes take effect immediately and apply only to calls made afterwards.

Model launch

seedream-5.0-flash launched · pricing

Model ID	Unit price (USD)
`seedream-5.0-flash`	Per image: $0.0162 (same at 1k / 1.5k / 2k; reference images free)

Pricing change

Video: SkyReels default-resolution pricing, source seconds on Wan 2.7 video editing, reference images on Grok Imagine Video

Model	Change
`skyreels-v4-fast` / `skyreels-v4-std`	Without `resolution`, calls are billed at the `1080p` tier (the resolution actually rendered); previously the `720p` tier. With a reference video, pass `480p` or `720p` explicitly
`wan2.7-videoedit`	The source video’s seconds are billed separately: `720p` $0.09 per second, `1080p` $0.135 per second, on top of output seconds (6-second minimum). This matches the “reference seconds + output seconds” rule announced on 2026-09-04
`grok-imagine-video`	Reference images are billed per image: $0.0018 per image

Changes take effect immediately and apply only to calls made afterwards.

Pricing change

Images: two economy lines cut; multi-image requests settle by images returned

Model	Change
`gemini-3.1-flash-lite-image-rev`	`1k` $0.015626 → $0.0140625 per image
`grok-imagine-image-2.0-rev`	$0.01875 → $0.016875 per image
`grok-imagine-image-2.0`	The $0.009 per-reference-image fee is now charged once per request, no longer repeated for each output image

When one request asks for several images (n above 1), it settles by the number of images actually returned; if fewer than n come back, the excess frozen quota is refunded when the task completes.Changes take effect immediately and apply only to calls made afterwards.

Model launch

2 models launched · pricingImage

Model ID	Unit price (USD)
`gpt-image-2.5-flare-rev`	Per image: 1k $0.0095625 · 2k $0.01575 · 4k $0.023625
`gpt-image-2.5-sunburst-rev`	Per image: 1k $0.0095625 · 2k $0.01575 · 4k $0.023625

Pricing change

Explicit caching enabled on qwen3.8-maxRequests can mark a reusable prefix with cache_control (previously rejected with 400). Writes are billed at the cache-write rate of $1.8567 and hits at the cache-read rate of $0.1854 (per 1M tokens); input, output and automatic cache-hit rates are unchanged.

Pricing change

Cache hits on 13 OpenAI models now billed at the cache-read rateInput tokens served from the automatic cache were previously billed at the regular input rate; they are now billed at the cache-read rate. Input and output rates are unchanged. Cache-read rates (per 1M tokens):

Model	Cache read
`gpt-5.4-mini`	$0.0675
`gpt-5.4-nano`	$0.018
`gpt-5.2`	$0.1575
`gpt-5.1`	$0.1125
`gpt-5`	$0.1125
`gpt-5-mini`	$0.0225
`gpt-5-nano`	$0.0045
`gpt-4.1`	$0.45
`gpt-4.1-mini`	$0.09
`gpt-4.1-nano`	$0.0225
`o4-mini`	$0.2475
`o3-mini`	$0.495
`o1`	$6.75

The number of cached tokens is reported in usage.prompt_tokens_details.cached_tokens.

Model launch

10 models launched · pricingText

Model ID	Price (USD)
`claude-opus-5-5`	input $3.6 · output $18 · cache read $0.18 · cache write $4.5 · cache write (1h) $7.2 (per 1M tokens)
`glm-5.3`	input $1.26 · output $3.96 · cache read $0.234 (per 1M tokens)
`gpt-6-luna`	input $0.09 · output $0.45 · cache read $0.009 · cache write $0.1125; above 272K input tokens: input $0.18 / output $0.675 (per 1M tokens)
`gpt-6-sol`	input $1.8 · output $9 · cache read $0.18 · cache write $2.25; above 272K input tokens: input $3.6 / output $15.75 (per 1M tokens)
`grok-4.7`	input $1.8 · output $5.4 · cache read $0.45; above 200K input tokens: input $3.6 / output $10.8 (per 1M tokens)
`qwen3.7-flash`	input $0.0252 · output $0.099 · cache read $0.0054 · cache write $0.0306; above 32K input tokens: input $0.0747 / output $0.297 (per 1M tokens)
`qwen3.8-2.4t-a95b`	input $1.485 · output $4.4559 · cache read $0.1854 · cache write $1.8567 (per 1M tokens)
`qwen3.8-27b`	input $0.3816 · output $1.5264 · cache read $0.0765 · cache write $0.477 (per 1M tokens)
`qwen3.8-flash`	input $0.1017 · output $0.3438 · cache read $0.0126 · cache write $0.1593 (per 1M tokens)
`qwen3.8-max-0902`	input $1.485 · output $4.4559 · cache read $0.1854 · cache write $1.8567 (per 1M tokens)

Pricing change

Three DeepSeek V4 models repriced; the schedule drops from three tiers to twoThe schedule now matches the upstream settlement windows: daily 00:00–14:00 UTC is the peak tier and 14:00–24:00 UTC is the off-peak tier (Beijing time is UTC+8). The earlier split — weekdays 01:00–04:00 and 06:00–10:00 at peak, everything else cheaper — is no longer used. Rates per window (per 1M tokens, input / cache read / output):

Model	Daily 00:00–14:00	Daily 14:00–24:00
`deepseek-v4-pro`	$1.1314 / $0.2263 / $3.3943	$0.5657 / $0.1131 / $1.6971
`deepseek-v4.1-flash`	$0.2514 / $0.0251 / $1.0057	$0.1257 / $0.0126 / $0.5029
`deepseek-v4-flash`	$0.3771 / $0.0754 / $1.1314	$0.1886 / $0.0377 / $0.5657

Every window goes up for all three models. Cache-hit rates rise the most, but stay well below the input rate of the same window. The full schedule is price_config.text_schedule on GET /v1/models, and it is the basis for settlement.

Pricing change

DeepSeek V4 Flash lowered to the new official priceDeepSeek has retired V4-Flash: deepseek-v4-flash requests are now served by V4.1-Flash and billed at the Flash price, so its time-of-day rates go down (per 1M tokens, input / cache read / output):

Window (UTC)	Change
Weekdays 01:00–04:00 and 06:00–10:00	$0.396 / $0.014 / $1.188 → $0.30 / $0.006 / $1.08
Weekdays 00:00–01:00, 04:00–06:00 and 10:00–14:00; weekends 00:00–14:00	$0.22 / $0.007 / $0.66 → $0.15 / $0.003 / $0.60
All other hours	$0.198 / $0.007 / $0.594 → $0.15 / $0.003 / $0.54

deepseek-v4-pro is unchanged. The newly added deepseek-v4.1-flash is also priced by time of day; the full schedule is price_config.text_schedule on GET /v1/models.

Model launch

3 models launched · pricingText

Model ID	Price (USD)
`claude-fable-5-1`	input $9 · output $45 · cache read $0.225 · cache write $11.25 · cache write (1h) $18 (per 1M tokens)
`gemini-3.8-flash`	input $0.675 · output $3.375 · cache read $0.0675 (per 1M tokens)
`gpt-6-astra`	input $9 · output $45 · cache read $0.9 · cache write $11.25; above 272K input tokens: input $18 / output $67.5 (per 1M tokens)

Pricing change

Two audio prices cut; Kling 3.0 Turbo and Nano Banana Economy re-based on official list prices

Model	Change
`whisper-1`	$0.03 → $0.0271 per minute (re-based on actual supplier cost; a cut)
`gpt-4o-mini-tts`	$13.5 → $2.25 per 1M characters (the upstream settles in text tokens; re-based on actual cost; a cut)
`kling-3.0-turbo`	720p $0.112 → $0.1287 /s, 1080p $0.14 → $0.1611 /s (aligned to APIMart list price × 0.9)
`gemini-3-pro-image-rev`	`1k` / `2k` $0.0375 → $0.03375 per image; `4k` added at $0.045
`gemini-3.1-flash-image-rev`	Re-based on list price × 0.9: `0.5k` / `1k` $0.016875, `2k` $0.0225, `4k` $0.028125 per image
`gemini-3.1-flash-lite-image-rev`	`1k` $0.015626 per image

Effective immediately; only calls made after the change are affected.

Pricing change

Reference-video billing realigned per vendor; gpt-image default quality; FLUX.2 reference images by MP tierCalls carrying a reference video used to be approximated as “output seconds × 2× the unit price”, which did not match how the upstreams actually charge. They are now billed per vendor, in four shapes:

Shape	Applies to	Formula
Reference seconds counted, tier switched	Seedance 2.x, Wan 2.7 R2V / VideoEdit	(reference-video seconds + output seconds) × the video-reference rate
Reference seconds counted, same tier	MiniMax H3 / H3-Max / H3-Regeneration	(reference-video seconds + output seconds) × the same tier’s rate
Tier switched, reference seconds free	Kling Omni / Video-O1, SkyReels V4	output seconds × the video-reference rate
Reference seconds only, no clip price	Omni-Flash-Ext	reference-video seconds × a per-second rate ($0.09 at 720p / 1080p, $0.27 at 4K); the clip price is not charged

For Kling’s two *-motion-control models the reference video carries no separate charge; output duration equals the reference video’s duration, and 720P bills at the standard tier, 1080P at the professional tier.Price cuts in this batch: seedance-2.0 family with a reference video (480p $0.082 → $0.045 per second, other tiers reduced proportionally — the video-reference rate is below the text-to-video rate) and Kling motion-control at 720P.Two more changes:

gpt-image-2 without quality now generates and bills at low instead of high (gpt-image-1 / gpt-image-1.5 use medium), matching OpenAI’s own defaults — this is a price cut. Passing auto explicitly still bills at the top tier.
FLUX.2 reference images are charged by MP tier: with exactly one reference image the surcharge follows that image’s actual MP tier (1–4 MP); with two or more, each is charged at the 1 MP tier.

Effective immediately, for calls made after the change. Live prices are in Explore and the console’s Model Market; POST https://api.qingbo.ai/waveapi/quote with video_urls / image_urls returns an estimate.

Pricing change

whisper-1 price changewhisper-1: $0.0054 → $0.03 per minute (billed per second, rounded up). The upstream bills whisper-1 above OpenAI’s list price and we realign to cost; for cheaper transcription use gpt-4o-mini-transcribe ($0.0027/min) or gpt-4o-transcribe ($0.0054/min), both with better accuracy. Effective immediately.

Pricing change

FLUX.2 realigned to megapixel tiers; Seedream 5.0 Lite price changeFLUX.2 is billed upstream by megapixel (MP) tier. 1K (1024×1024 ≈ 1.05 MP) actually falls in the 2 MP tier; it had been priced at the 1 MP tier, and is realigned today:

Model	`1K`	`2K`	Reference image / each
`flux-2-pro`	$0.027 → $0.0405	$0.0675 (unchanged)	$0.0135 → $0.027
`flux-2-flex`	$0.045 → $0.09	$0.18 (unchanged)	$0.045 → $0.09
`flux-2-max`	$0.063 → $0.09	$0.144 (unchanged)	$0.027 → $0.054

seedream-5.0-lite: $0.028286 → $0.035 per image (no upstream discount on the list price; adjusted per our pricing rule).seedream-4.5 no longer offers 1K (upstream only bills 2K / 4K); imagen-4.0 is delisted (upstream no longer serves it). Effective immediately.

Model launch

2 models launched · pricingVideo

Model ID	Price (USD)
`gemini-omni-1.1-flash`	Per second: 360p $0.0333 · 720p $0.099 · 1080p $0.1485 · 4k $0.297/s
`omni-flash-ext`	Per clip: 360p 4s $0.16875 · 360p 6s $0.19687 · 360p 8s $0.225 · 360p 10s $0.25312 · 720p 4s $0.28125 · 720p 6s $0.3375 · 720p 8s $0.39375 · 720p 10s $0.45; reference video per second 360p $0.045 · 720p $0.09 · 1080p $0.09 · 4k $0.27

Billing change

Image models: reference images now billed per imageFor the models below, reference images (image_urls) are now charged per image, matching the vendors’ official pricing (previously some models charged nothing and some charged an upper-bound tier):

Model	Per reference image
`seedream-5.0-pro`	$0.002571
`qwen-image-3.0` / `qwen-image-3.0-pro`	$0.002571
`grok-imagine-image`	$0.0018
`grok-imagine-image-quality` / `grok-imagine-image-2.0`	$0.009
`flux-2-pro` / `flux-2-max` / `flux-2-flex`	$0.0135 / $0.027 / $0.045 (up to 8 reference images now accepted)
`gpt-image-2` / `gpt-image-1.5`	$0.0025 (gpt-image-2 previously used 1–3 / 4+ image tiers; now linear)
`gpt-image-1`	$0.0031

Cost per call = output tier price + number of reference images × unit price; calls without reference images are unaffected. Seedream 4.x / 5.0-lite, Qwen Image 2.0, Wan 2.7 Image and the Nano Banana family do not charge for reference images. Live prices are on Explore and in the console “Model Market”; POST https://api.qingbo.ai/waveapi/quote with image_urls returns an estimate.

Pricing change

Nano Banana Pro / Nano Banana 2 standard tiers — price cutThe upstream list price of the standard (-rev) tiers dropped; ours follow:

Model	1K / 2K	4K
`gemini-3-pro-image-preview-rev` (Nano Banana Pro, standard)	$0.045 → $0.03375	$0.073125 → $0.045
`gemini-3.1-flash-image-preview-rev` (Nano Banana 2, standard)	1K $0.03375 → $0.016875, 2K $0.045 → $0.0225	$0.0675 → $0.028125

Official direct tiers (without -rev) are unchanged. Effective immediately.

Pricing change

Text-model prices re-aligned to vendors’ official list prices; long-context tiers addedWhile onboarding new text models we cross-checked existing text groups against each vendor’s official pricing page:

Model	Change
`glm-5.1` / `glm-4.6`	Reference prices had been taken from an upstream conversion that was too low; now priced from Z.ai’s official USD list (raised)
`minimax-m2.1`	Aligned to MiniMax’s official legacy tier (slight increase)
`kimi-k2`	Input price had been above the official reference; lowered to 0.9 × reference
`deepseek-r1-0528` / `deepseek-v3.2` / `deepseek-v3.2-exp`	Output prices unified at 0.9 × reference (r1-0528 lowered; the two v3.2 tiers slightly raised)
GPT-5.6 Sol / Terra / Luna, GPT-5.5, GPT-5.4, GPT-5.4 Pro	Long-context tier added above 272K input (matching OpenAI; previously such requests were billed at the standard rate)
Gemini 3.1 Pro / 3 Pro / 2.5 Pro	Long-context tier added above 200K input (matching Google)
`kimi-k3` / `claude-sonnet-5`	Cached-input price added (cache hits billed at the lower rate)

Changes take effect immediately and apply only to calls made afterwards. Live unit prices are shown on Explore and in the console “Model Market”.

Pricing change

Existing models re-aligned to vendors’ official list prices (both up and down)While onboarding new models we cross-checked existing ones against each vendor’s official pricing page. The following unit prices have been synced:

Model	Change
`kimi-k3`	Output price lowered to Moonshot’s official list price
`gpt-image-2`	Requests without `quality` are now billed at the `high` tier (previously a lower tier; now matches OpenAI’s default)
`seedream-5.0` / `seedream-5.0-lite`	All tiers synced to Volcengine’s list price; `1.5K` tier added for 5.0-lite
`seedance-2.0` / `2.0-fast` / `2.0-face` / `2.0-fast-face` / `2.0-mini` / `2.5` / `1.5-pro`	Image-to-video now same price as text-to-video (lowered); reference-video requests billed at 2× (raised); tiers previously set from promotional prices moved back to list price
`wan2.6` / `wan2.7` / `wan2.7-r2v` / `wan2.7-videoedit` / `wan2.5-preview`	All tiers synced to Alibaba Cloud Model Studio list prices; video edit / reference-to-video billed at 2×
`minimax-h3`	All tiers synced to MiniMax’s official price list

Changes take effect immediately and apply only to calls made afterwards. Live unit prices are shown in the Models overview and the console “Model Market”.

Model launch

midjourney launched · pricing

Model ID	Price (USD)
`midjourney`	Per image: imagine $0.05067 · upscale $0.06192 · variation $0.06192 · high-variation $0.06192 · low-variation $0.06192 · reroll $0.06192 · other actions $0.05067–$0.06192

Model launch

suno launched · pricing

Model ID	Price (USD)
`suno`	Per call: generate $0.05625 · extend $0.05625 · cover $0.05625 · lyrics $0.009 · stems $0.1125 · inspo $0.0765 · download $0.0018 · other actions $0.0009–$0.27

Model launch

gemini-3.7-flash launched · pricing

Model ID	Price (USD)
`gemini-3.7-flash`	input $0.675 · output $3.375 · cache read $0.0675 (per 1M tokens)

Model launch

grok-4.6 launched · pricing

Model ID	Price (USD)
`grok-4.6`	input $1.8 · output $5.4 · cache read $0.45; above 200K input tokens: input $3.6 / output $10.8 (per 1M tokens)

Model launch

2 models launched · pricingImage

Model ID	Price (USD)
`grok-imagine-image-2.0`	Per image: 1k $0.054 · 2k $0.072; reference image $0.009 each
`grok-imagine-image-2.0-rev`	Per image: $0.01688

Model launch

wan3.0-video launched · pricing

Model ID	Price (USD)
`wan3.0-video`	Per second: 480p $0.045 · 720p $0.09 · 1080p $0.18/s

Model launch

3 models launched · pricingImage

Model ID	Price (USD)
`qwen-image-3.0`	Per image: 1k $0.02228 · 2k $0.02228; reference image $0.00248 each

Video

Model ID	Price (USD)
`flux-3-video`	Per second: with reference video fhd $0.477 · hd $0.369/s
`flux-3-video-draft`	Per second: with reference video hd $0.108/s

Model launch

qwen3.8-max launched · pricing

Model ID	Price (USD)
`qwen3.8-max`	input $1.485 · output $4.4559 · cache read $0.1854 · cache write $1.8567 (per 1M tokens)

Model launch

4 models launched · pricingVideo

Model ID	Price (USD)
`minimax-h3`	Per second: 768p $0.072 · 2k $0.117/s; with reference video 768p $0.072 · 2k $0.117/s; reference video per second 768p $0.072 · 2k $0.117; reference images beyond 5: $0.036 each
`minimax-h3-context-ir` (retired)	input $0.87 · output $3.45 (per 1M tokens)
`minimax-h3-regeneration`	Per second: 2k $0.045/s; with reference video 2k $0.045/s; reference video per second 2k $0.045; reference images beyond 5: $0.0225 each
`seedance-2.5`	Per second: 480p $0.103 · 720p $0.231 · 1080p $0.409/s; with reference video 480p $0.072 · 720p $0.1458 · 1080p $0.25866/s; reference video per second 480p $0.072 · 720p $0.1458 · 1080p $0.25866

Model launch

claude-opus-5 launched · pricing

Model ID	Price (USD)
`claude-opus-5`	input $4.5 · output $22.5 · cache read $0.45 · cache write $5.625 · cache write (1h) $9 (per 1M tokens)

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`gemini-3.5-flash-lite`	input $0.27 · output $2.25 (per 1M tokens)
`gemini-3.6-flash`	input $1.35 · output $6.75 · cache read $0.135 (per 1M tokens)

Model launch

qwen-image-3.0-pro launched · pricing

Model ID	Price (USD)
`qwen-image-3.0-pro`	Per image: 1k $0.03094 · 2k $0.06188; reference image $0.00248 each

Model launch

kimi-k3 launched · pricing

Model ID	Price (USD)
`kimi-k3`	input $2.7 · output $13.5 · cache read $0.27 (per 1M tokens)

Model launch

3 models launched · pricingText

Model ID	Price (USD)
`gpt-5.6-luna`	input $0.18 · output $1.08 · cache read $0.018 · cache write $0.225; above 272K input tokens: input $0.36 / output $1.62 (per 1M tokens)
`gpt-5.6-sol`	input $3.6 · output $18 · cache read $0.36 · cache write $4.5; above 272K input tokens: input $7.2 / output $27 (per 1M tokens)
`gpt-5.6-terra`	input $1.8 · output $10.8 · cache read $0.18 · cache write $2.25; above 272K input tokens: input $3.6 / output $16.2 (per 1M tokens)

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`grok-4.5`	input $1.8 · output $5.4; above 200K input tokens: input $3.6 / output $10.8 (per 1M tokens)

Image

Model ID	Price (USD)
`seedream-5.0-pro`	Per image: 1k $0.0405 · 2k $0.081

Pricing change

Pricing synced to official for some image / video modelsFollowing upstream official price changes, unit prices for the following models have been synced:

Model	Change
`gpt-image-2`	1K / 2K / 4K tiers increased
`gemini-3-pro-image-preview`	4K tier increased
`veo3.1-fast` / `veo3.1-lite` / `veo3.1-quality`	All resolution tiers synced to official (both increases and decreases)
`veo3.1-fast`	All tiers decreased
`omni-flash-ext`	Now billed by resolution × duration — see “Model Updates”

Effective immediately; applies only to calls made after the change. Current prices are shown in the Models overview and the console “Model Market”.

Model launch

4 models launched · pricingText

Model ID	Price (USD)
`claude-sonnet-5`	input $1.8 · output $9 · cache read $0.18 · cache write $2.25 · cache write (1h) $3.6 (per 1M tokens)

Image

Model ID	Price (USD)
`gemini-3.1-flash-lite-image`	By actual token usage: image input $0.225 · image output $27 · text input $0.225 · text output $1.35 (per 1M tokens)
`gemini-3.1-flash-lite-image-rev`	Per image: 1k $0.01406

Video

Model ID	Price (USD)
`gemini-omni-flash-preview`	Per second: 720p $0.099/s

Model launch

happyhorse-1.1 launched · pricing

Model ID	Price (USD)
`happyhorse-1.1`	Per second: 720p $0.14625 · 1080p $0.1935/s

Pricing change

Seedance 2.0 pricing synced to official

seedance-2.0 unit prices are now synced to official rates, and the new 4K tier is live for billing. Current prices are shown in the Models overview and the console “Model Market”.

Model launch

kling-3.0-turbo launched · pricing

Model ID	Price (USD)
`kling-3.0-turbo`	Per second: 720p $0.1287 · 1080p $0.1611/s

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`glm-5.2`	input $1.26 · output $3.96 (per 1M tokens)

Video

Model ID	Price (USD)
`seedance-2.0-mini`	Per second: 480p $0.0144 · 720p $0.0288/s; with reference video 480p $0.0072 · 720p $0.01557/s; reference video per second 480p $0.0072 · 720p $0.01557

Model launch

kimi-k2.7-code-highspeed launched · pricing

Model ID	Price (USD)
`kimi-k2.7-code-highspeed`	input $1.71 · output $7.2 (per 1M tokens)

Model launch

kimi-k2.7-code launched · pricing

Model ID	Price (USD)
`kimi-k2.7-code`	input $0.855 · output $3.6 (per 1M tokens)

Pricing change

Pricing synced for some video models

Unit prices for some video models were synced to official rates. Current prices are shown in the Models overview and the console “Model Market”.

Model launch

claude-fable-5 launched · pricing

Model ID	Price (USD)
`claude-fable-5`	input $9 · output $45 · cache read $0.9 · cache write $11.25 · cache write (1h) $18 (per 1M tokens)

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`minimax-m3`	input $0.27 · output $1.08 (per 1M tokens)
`qwen3.7-plus`	input $0.2208 · output $0.8808; above 256K input tokens: input $0.59472 / output $2.37672 (per 1M tokens)

Model launch

4 models launched · pricingImage

Model ID	Price (USD)
`grok-imagine-1.5-edit-rev`	Per image: $0.01688
`grok-imagine-1.5-rev`	Per image: $0.01688

Video

Model ID	Price (USD)
`grok-imagine-video-1.5`	Per second: 480p $0.072 · 720p $0.126 · 1080p $0.225/s; reference images beyond 0: $0.009 each
`grok-imagine-video-1.5-rev`	Per second: 480p $0.01148 · 720p $0.02151/s

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`grok-build-0.1`	input $0.9 · output $1.8; above 200K input tokens: input $1.8 / output $3.6 (per 1M tokens)
`step-3.7-flash`	input $0.18 · output $1.035 (per 1M tokens)

Model launch

claude-opus-4-8 launched · pricing

Model ID	Price (USD)
`claude-opus-4-8`	input $4.5 · output $22.5 (per 1M tokens)

Model launch

qwen3.7-max launched · pricing

Model ID	Price (USD)
`qwen3.7-max`	input $1.485 · output $4.4559 (per 1M tokens)

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`gemini-3.5-flash`	input $1.35 · output $8.1 (per 1M tokens)
`gemini-flash-latest` (retired)	input $1.35 · output $8.1 (per 1M tokens)

Model launch

grok-4.3 launched · pricing

Model ID	Price (USD)
`grok-4.3`	input $1.125 · output $2.25; above 200K input tokens: input $2.25 / output $4.5 (per 1M tokens)

Model launch

happyhorse-1.0 launched · pricing

Model ID	Price (USD)
`happyhorse-1.0`	Per second: 720p $0.14625 · 1080p $0.25875/s

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`deepseek-v4-flash`	(per 1M tokens)
`deepseek-v4-pro`	(per 1M tokens)

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`gpt-5.5`	input $4.5 · output $27 · cache read $0.45; above 272K input tokens: input $9 / output $40.5 (per 1M tokens)
`mimo-v2.5-pro`	input $0.3915 · output $0.783 (per 1M tokens)

Model launch

2 models launched · pricingImage

Model ID	Price (USD)
`gpt-image-2`	By actual token usage: cached image input $1.8 · cached text input $1.125 · image input $7.2 · image output $27 · text input $4.5 (per 1M tokens)
`gpt-image-2-rev`	Per image: 1k $0.00956 · 2k $0.01575 · 4k $0.02363

Model launch

kimi-k2.6 launched · pricing

Model ID	Price (USD)
`kimi-k2.6`	input $0.855 · output $3.6 (per 1M tokens)

Model launch

flow-music launched · pricing

Model ID	Price (USD)
`flow-music`	Per call: generate $0.0675 · extend $0.0675 · cover $0.0675 · lyrics $0.0225 · stems $0.0675 · other actions $0.01125–$0.0675

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`claude-opus-4-7`	input $4.5 · output $22.5 (per 1M tokens)
`qwen3.6-flash`	input $0.1485 · output $0.891; above 256K input tokens: input $0.594 / output $3.5649 (per 1M tokens)

Model launch

2 models launched · pricingVideo

Model ID	Price (USD)
`vidu-q3`	Per second: 540p $0.045 · 720p $0.09 · 1080p $0.1125/s
`vidu-q3-mix`	Per second: 720p $0.1125 · 1080p $0.135/s

Model launch

glm-5.1 launched · pricing

Model ID	Price (USD)
`glm-5.1`	input $1.26 · output $3.96 (per 1M tokens)

Model launch

4 models launched · pricingImage

Model ID	Price (USD)
`grok-imagine-image-quality`	Per image: 1k $0.045 · 2k $0.063; reference image $0.009 each

Video

Model ID	Price (USD)
`wan2.7`	Per second: 720p $0.09 · 1080p $0.135/s; with reference video 720p $0.09 · 1080p $0.135/s; reference video per second 720p $0.09 · 1080p $0.135
`wan2.7-r2v`	Per second: 720p $0.09 · 1080p $0.135/s; with reference video 720p $0.09 · 1080p $0.135/s; reference video per second 720p $0.09 · 1080p $0.135
`wan2.7-videoedit`	Per second: 720p $0.09 · 1080p $0.135/s

Model launch

qwen3.6-plus launched · pricing

Model ID	Price (USD)
`qwen3.6-plus`	input $0.2484 · output $1.4859; above 256K input tokens: input $0.9909 / output $5.9418 (per 1M tokens)

Model launch

2 models launched · pricingImage

Model ID	Price (USD)
`wan2.7-image`	Per image: $0.027
`wan2.7-image-pro`	Per image: $0.0675

Model launch

2 models launched · pricingVideo

Model ID	Price (USD)
`pixverse-v6`	Per second: 360p $0.018 · 540p $0.027 · 720p $0.036 · 1080p $0.072/s; with audio 360p $0.027 · 540p $0.036 · 720p $0.045 · 1080p $0.09/s
`veo3.1-lite`	Per clip: 720p $0.07875 · 1080p $0.07875

Model launch

2 models launched · pricingVideo

Model ID	Price (USD)
`skyreels-v4-fast`	Per second: 480p $0.072 · 720p $0.099 · 1080p $0.2475/s; with reference video 480p $0.135 · 720p $0.18 · 1080p $0.45/s
`skyreels-v4-std`	Per second: 480p $0.099 · 720p $0.126 · 1080p $0.315/s; with reference video 480p $0.162 · 720p $0.225 · 1080p $0.5625/s

Model launch

minimax-m2.7 launched · pricing

Model ID	Price (USD)
`minimax-m2.7`	input $0.27 · output $1.08 (per 1M tokens)

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`gpt-5.4-mini`	input $0.675 · output $4.05 · cache read $0.0675 (per 1M tokens)
`gpt-5.4-nano`	input $0.18 · output $1.125 · cache read $0.018 (per 1M tokens)

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`grok-4.20-0309-non-reasoning`	input $1.125 · output $2.25; above 200K input tokens: input $2.25 / output $4.5 (per 1M tokens)
`grok-4.20-0309-reasoning`	input $1.125 · output $2.25; above 200K input tokens: input $2.25 / output $4.5 (per 1M tokens)

Model launch

3 models launched · pricingText

Model ID	Price (USD)
`gpt-5.4`	input $2.25 · output $13.5 · cache read $0.225; above 272K input tokens: input $4.5 / output $20.25 (per 1M tokens)
`gpt-5.4-pro`	input $27 · output $162; above 272K input tokens: input $54 / output $243 (per 1M tokens)

Video

Model ID	Price (USD)
`kling-v3-motion-control`	Per second: 720p $0.1134 · 1080p $0.1512/s

Model launch

gemini-3.1-flash-lite-preview launched · pricing

Model ID	Price (USD)
`gemini-3.1-flash-lite-preview` (retired)	input $0.225 · output $1.35 (per 1M tokens)

Model launch

2 models launched · pricingImage

Model ID	Price (USD)
`gemini-3.1-flash-image`	By actual token usage: image input $0.45 · image output $54 · text input $0.45 · text output $2.7 (per 1M tokens)
`gemini-3.1-flash-image-rev`	Per image: 1k $0.01688 · 2k $0.0225 · 4k $0.02813

Model launch

gpt-5.3-codex launched · pricing

Model ID	Price (USD)
`gpt-5.3-codex`	input $1.575 · output $12.6 (per 1M tokens)

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`gemini-3.1-pro-preview`	input $1.8 · output $10.8; above 200K input tokens: input $3.6 / output $16.2 (per 1M tokens)
`gemini-3.1-pro-preview-customtools` (retired)	input $1.8 · output $10.8; above 200K input tokens: input $3.6 / output $16.2 (per 1M tokens)

Model launch

claude-sonnet-4-6 launched · pricing

Model ID	Price (USD)
`claude-sonnet-4-6`	input $2.7 · output $13.5 (per 1M tokens)

Model launch

seedream-5.0-lite launched · pricing

Model ID	Price (USD)
`seedream-5.0-lite`	Per image: 2k $0.0315 · 3k $0.0315 · 4k $0.0315

Model launch

6 models launched · pricingText

Model ID	Price (USD)
`glm-5`	input $0.9 · output $2.88 (per 1M tokens)
`minimax-m2.5`	input $0.27 · output $1.08 (per 1M tokens)

Video

Model ID	Price (USD)
`seedance-2.0`	Per second: 480p $0.07 · 720p $0.15 · 1080p $0.37 · 4k $0.78/s; with reference video 480p $0.045 · 720p $0.09657 · 1080p $0.24264 · 4k $0.49986/s; reference video per second 480p $0.045 · 720p $0.09657 · 1080p $0.24264 · 4k $0.49986
`seedance-2.0-face` (retired 2026-09-08)	Per second: 480p $0.07 · 720p $0.15 · 1080p $0.37/s
`seedance-2.0-fast`	Per second: 480p $0.0405 · 720p $0.09/s; with reference video 480p $0.02664 · 720p $0.05769/s; reference video per second 480p $0.02664 · 720p $0.05769
`seedance-2.0-fast-face` (retired 2026-09-08)	Per second: 480p $0.0405 · 720p $0.09/s

Model launch

vidu-q3-turbo launched · pricing

Model ID	Price (USD)
`vidu-q3-turbo`	Per second: 540p $0.036 · 720p $0.054 · 1080p $0.063/s

Model launch

2 models launched · pricingImage

Model ID	Price (USD)
`qwen-image-2.0`	Per image: $0.0225
`qwen-image-2.0-pro`	Per image: $0.05625

Model launch

3 models launched · pricingText

Model ID	Price (USD)
`claude-opus-4-6`	input $4.5 · output $22.5 (per 1M tokens)

Video

Model ID	Price (USD)
`kling-v3`	Per second: 720p $0.0756 · 1080p $0.1008/s; with audio 720p $0.1134 · 1080p $0.1512/s
`kling-v3-omni`	Per second: 720p $0.0756 · 1080p $0.1008/s; with audio 720p $0.1008 · 1080p $0.126/s; with reference video 720p $0.1134 · 1080p $0.1512/s

Model launch

3 models launched · pricingImage

Model ID	Price (USD)
`grok-imagine-1.0-edit-rev` (retired 2026-09-08)	Per image: $0.01688
`grok-imagine-1.0-rev` (retired 2026-09-08)	Per image: $0.01688

Video

Model ID	Price (USD)
`grok-imagine-1.0-video-rev` (retired 2026-09-23)	Per second: 480p $0.01148 · 720p $0.02151/s

Model launch

2 models launched · pricingImage

Model ID	Price (USD)
`grok-imagine-image`	Per image: 1k $0.018 · 2k $0.018; reference image $0.0018 each

Video

Model ID	Price (USD)
`grok-imagine-video`	Per second: 480p $0.045 · 720p $0.063/s

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`kimi-k2.5` (retired)	input $0.54 · output $2.7 (per 1M tokens)

Video

Model ID	Price (USD)
`vidu-q3-pro`	Per second: 540p $0.063 · 720p $0.135 · 1080p $0.144/s

Model launch

wan2.6-i2v-flash launched · pricing

Model ID	Price (USD)
`wan2.6-i2v-flash`	Per second: 720p $0.0225 · 1080p $0.03375/s; with audio 720p $0.045 · 1080p $0.0675/s

Model launch

gpt-5.2-codex launched · pricing

Model ID	Price (USD)
`gpt-5.2-codex` (retired)	input $1.575 · output $12.6 (per 1M tokens)

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`glm-4.7`	input $0.54 · output $1.98 (per 1M tokens)
`minimax-m2.1`	input $0.27 · output $1.08 (per 1M tokens)

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`gemini-3-flash-preview`	input $0.45 · output $2.7 (per 1M tokens)

Video

Model ID	Price (USD)
`kling-v2-6-motion-control`	Per second: 720p $0.063 · 1080p $0.1008/s

Model launch

5 models launched · pricingImage

Model ID	Price (USD)
`flux-2-max`	Per image: $0.09
`gpt-image-1.5`	By actual token usage: cached image input $1.8 · cached text input $1.125 · image input $7.2 · image output $28.8 · text input $4.5 · text output $9 (per 1M tokens)

Video

Model ID	Price (USD)
`seedance-1.5-pro`	Per second: 480p $0.0216 · 720p $0.0468 · 1080p $0.116/s; with audio 480p $0.0216 · 720p $0.0468 · 1080p $0.116/s
`wan2.6`	Per second: 1080p $0.0945/s
`wan2.6-i2v`	Per second: 720p $0.0747 · 1080p $0.1233/s

Model launch

3 models launched · pricingText

Model ID	Price (USD)
`gpt-5.2`	input $1.575 · output $12.6 · cache read $0.1575 (per 1M tokens)
`gpt-5.2-chat-latest` (retired)	input $1.575 · output $12.6 (per 1M tokens)
`gpt-5.2-pro`	input $18.9 · output $151.2 (per 1M tokens)

Model launch

gpt-5.1-codex-max launched · pricing

Model ID	Price (USD)
`gpt-5.1-codex-max` (retired)	input $1.125 · output $9 (per 1M tokens)

Model launch

2 models launched · pricingImage

Model ID	Price (USD)
`seedream-4.5`	Per image: 2k $0.036 · 4k $0.036

Video

Model ID	Price (USD)
`kling-v2-6`	Per second: 720p $0.0378 · 1080p $0.063/s; with audio 1080p $0.126/s

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`deepseek-v3.2`	input $0.252 · output $0.378 (per 1M tokens)

Video

Model ID	Price (USD)
`kling-video-o1`	Per second: 720p $0.0756 · 1080p $0.1008/s; with reference video 720p $0.1134 · 1080p $0.1512/s

Model launch

z-image-turbo launched · pricing

Model ID	Price (USD)
`z-image-turbo`	Per image: $0.01125

Model launch

2 models launched · pricingImage

Model ID	Price (USD)
`flux-2-flex`	Per image: $0.09
`flux-2-pro`	Per image: $0.0405

Model launch

claude-opus-4-5 launched · pricing

Model ID	Price (USD)
`claude-opus-4-5`	input $4.5 · output $22.5 (per 1M tokens)

Model launch

2 models launched · pricingImage

Model ID	Price (USD)
`gemini-3-pro-image`	By actual token usage: image input $1.8 · image output $108 · text input $1.8 · text output $10.8 (per 1M tokens)
`gemini-3-pro-image-rev`	Per image: 1k $0.03375 · 2k $0.03375 · 4k $0.045

Model launch

gemini-3-pro-preview launched · pricing

Model ID	Price (USD)
`gemini-3-pro-preview` (retired)	input $1.8 · output $10.8; above 200K input tokens: input $3.6 / output $16.2 (per 1M tokens)

Model launch

5 models launched · pricingText

Model ID	Price (USD)
`gpt-5.1`	input $1.125 · output $9 · cache read $0.1125 (per 1M tokens)
`gpt-5.1-2025-11-13` (retired)	input $1.125 · output $9 (per 1M tokens)
`gpt-5.1-chat-latest` (retired)	input $1.125 · output $9 (per 1M tokens)
`gpt-5.1-codex` (retired)	input $1.125 · output $9 (per 1M tokens)
`gpt-5.1-codex-mini` (retired)	input $0.225 · output $1.8 (per 1M tokens)

Model launch

2 models launched · pricingVideo

Model ID	Price (USD)
`minimax-hailuo-2.3`	Per second: 768p $0.0549 · 1080p $0.081/s
`minimax-hailuo-2.3-fast`	Per second: 768p $0.0279 · 1080p $0.0477/s

Model launch

seedance-1.0-pro-fast launched · pricing

Model ID	Price (USD)
`seedance-1.0-pro-fast`	Per second: 480p $0.009 · 720p $0.02 · 1080p $0.0432/s

Model launch

deepseek-ocr launched · pricing

Model ID	Price (USD)
`deepseek-ocr` (retired)	input $0.0513 · output $0.23085 (per 1M tokens)

Model launch

5 models launched · pricingText

Model ID	Price (USD)
`claude-haiku-4-5`	input $0.9 · output $4.5 (per 1M tokens)

Video

Model ID	Price (USD)
`veo3.1-fast`	Per second: 720p $0.072 · 1080p $0.09/s; with audio 720p $0.09 · 1080p $0.108/s
`veo3.1-fast-rev`	Per clip: 720p $0.1575 · 1080p $0.1575
`veo3.1-quality`	Per second: 720p $0.18 · 1080p $0.18/s; with audio 720p $0.36 · 1080p $0.36/s
`veo3.1-quality-rev`	Per clip: 720p $1.125 · 1080p $1.125

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`gpt-5-search-api` (retired)	input $1.125 · output $9 (per 1M tokens)
`gpt-5-search-api-2025-10-14` (retired)	input $1.125 · output $9 (per 1M tokens)

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`gpt-5-pro`	input $13.5 · output $108 (per 1M tokens)
`gpt-5-pro-2025-10-06` (retired)	input $13.5 · output $108 (per 1M tokens)

Model launch

4 models launched · pricingText

Model ID	Price (USD)
`glm-4.6`	input $0.54 · output $1.98 (per 1M tokens)

Video

Model ID	Price (USD)
`sora-2` (retired 2026-09-24)	Per second: 720p $0.09/s
`sora-2-preview` (retired 2026-09-03)	Per second: 720p $0.09/s
`sora-2-pro` (retired 2026-09-24)	Per second: 720p $0.27 · 1024p $0.45 · 1080p $0.63/s

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`claude-sonnet-4-5`	input $2.7 · output $13.5 (per 1M tokens)
`deepseek-v3.2-exp`	input $0.252 · output $0.378 (per 1M tokens)

Model launch

wan2.5-preview launched · pricing

Model ID	Price (USD)
`wan2.5-preview`	Per second: 480p $0.0378 · 720p $0.0747 · 1080p $0.1233/s

Model launch

gpt-5-codex launched · pricing

Model ID	Price (USD)
`gpt-5-codex` (retired)	input $1.125 · output $9 (per 1M tokens)

Model launch

deepseek-v3.1-terminus launched · pricing

Model ID	Price (USD)
`deepseek-v3.1-terminus`	input $0.504 · output $1.512 (per 1M tokens)

Model launch

seedream-4.0 launched · pricing

Model ID	Price (USD)
`seedream-4.0`	Per image: 1k $0.027 · 2k $0.027 · 4k $0.027

Model launch

2 models launched · pricingImage

Model ID	Price (USD)
`gemini-2.5-flash-image-preview` (retired)	Per image: $0.01406
`gemini-2.5-flash-image-preview-rev`	Per image: $0.01406

Model launch

imagen-4.0 launched · pricing

Model ID	Price (USD)
`imagen-4.0` (retired 2026-09-03)	Per image: $0.045

Model launch

7 models launched · pricingText

Model ID	Price (USD)
`gpt-5`	input $1.125 · output $9 · cache read $0.1125 (per 1M tokens)
`gpt-5-2025-08-07` (retired)	input $1.125 · output $9 (per 1M tokens)
`gpt-5-chat-latest` (retired)	input $1.125 · output $9 (per 1M tokens)
`gpt-5-mini`	input $0.225 · output $1.8 · cache read $0.0225 (per 1M tokens)
`gpt-5-mini-2025-08-07` (retired)	input $0.225 · output $1.8 (per 1M tokens)
`gpt-5-nano`	input $0.045 · output $0.36 · cache read $0.0045 (per 1M tokens)
`gpt-5-nano-2025-08-07` (retired)	input $0.045 · output $0.36 (per 1M tokens)

Model launch

gemini-2.5-flash-lite launched · pricing

Model ID	Price (USD)
`gemini-2.5-flash-lite`	input $0.09 · output $0.36 (per 1M tokens)

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`kimi-k2` (retired)	input $0.45 · output $1.89 (per 1M tokens)
`kimi-k2-instruct` (retired 2026-09-03)	input $0.45 · output $1.89 (per 1M tokens)

Model launch

minimax-hailuo-02 launched · pricing

Model ID	Price (USD)
`minimax-hailuo-02`	Per second: 512p $0.0117 · 768p $0.045 · 1080p $0.09/s

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`gemini-2.5-flash` (retired)	input $0.27 · output $2.25 (per 1M tokens)
`gemini-2.5-pro`	input $1.125 · output $9; above 200K input tokens: input $2.25 / output $13.5 (per 1M tokens)

Model launch

seedance-1.0-pro launched · pricing

Model ID	Price (USD)
`seedance-1.0-pro`	Per second: 480p $0.0216 · 720p $0.0468 · 1080p $0.1044/s

Model launch

o3-pro launched · pricing

Model ID	Price (USD)
`o3-pro`	input $18 · output $72 (per 1M tokens)

Model launch

2 models launched · pricingImage

Model ID	Price (USD)
`flux-kontext-max`	Per image: $0.072
`flux-kontext-pro`	Per image: $0.036

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`deepseek-r1-0528`	input $0.495 · output $1.971 (per 1M tokens)
`deepseek-r1-250528` (retired 2026-09-03)	input $0.504 · output $2.016 (per 1M tokens)

Model launch

gpt-image-1 launched · pricing

Model ID	Price (USD)
`gpt-image-1`	By actual token usage: cached image input $2.25 · cached text input $1.125 · image input $9 · image output $36 · text input $4.5 (per 1M tokens)

Model launch

4 models launched · pricingText

Model ID	Price (USD)
`o3`	input $1.8 · output $7.2 (per 1M tokens)
`o3-2025-04-16` (retired)	input $1.8 · output $7.2 (per 1M tokens)
`o4-mini`	input $0.99 · output $3.96 · cache read $0.2475 (per 1M tokens)
`o4-mini-2025-04-16` (retired)	input $0.99 · output $3.96 (per 1M tokens)

Model launch

6 models launched · pricingText

Model ID	Price (USD)
`gpt-4.1`	input $1.8 · output $7.2 · cache read $0.45 (per 1M tokens)
`gpt-4.1-2025-04-14` (retired)	input $1.8 · output $7.2 (per 1M tokens)
`gpt-4.1-mini`	input $0.36 · output $1.44 · cache read $0.09 (per 1M tokens)
`gpt-4.1-mini-2025-04-14` (retired)	input $0.36 · output $1.44 (per 1M tokens)
`gpt-4.1-nano`	input $0.09 · output $0.36 · cache read $0.0225 (per 1M tokens)
`gpt-4.1-nano-2025-04-14` (retired)	input $0.09 · output $0.36 (per 1M tokens)

Model launch

deepseek-v3-0324 launched · pricing

Model ID	Price (USD)
`deepseek-v3-0324` (retired 2026-09-03)	input $0.243 · output $0.99 (per 1M tokens)

Model launch

3 models launched · pricingAudio

Model ID	Price (USD)
`gpt-4o-mini-transcribe`	$0.0027 per minute of audio
`gpt-4o-mini-tts`	$2.25 per 1M characters of input text
`gpt-4o-transcribe`	$0.0054 per minute of audio

Model launch

o1-pro launched · pricing

Model ID	Price (USD)
`o1-pro` (retired)	input $135 · output $540 (per 1M tokens)

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`o3-mini`	input $0.99 · output $3.96 · cache read $0.495 (per 1M tokens)
`o3-mini-2025-01-31` (retired)	input $0.99 · output $3.96 (per 1M tokens)

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`o1`	input $13.5 · output $54 · cache read $6.75 (per 1M tokens)
`o1-2024-12-17` (retired)	input $13.5 · output $54 (per 1M tokens)

Model launch

gpt-4o-2024-11-20 launched · pricing

Model ID	Price (USD)
`gpt-4o-2024-11-20` (retired)	input $2.25 · output $9 (per 1M tokens)

Model launch

omni-moderation-latest launched · pricing

Model ID	Price (USD)
`omni-moderation-latest` (retired)	input $0.18 · output $0.18 (per 1M tokens)

Model launch

gpt-4o-2024-08-06 launched · pricing

Model ID	Price (USD)
`gpt-4o-2024-08-06` (retired 2026-09-03)	input $2.25 · output $9 (per 1M tokens)

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`gpt-4o-mini` (retired)	input $0.135 · output $0.54 (per 1M tokens)
`gpt-4o-mini-2024-07-18` (retired)	input $0.135 · output $0.54 (per 1M tokens)

Model launch

2 models launched · pricingText

Model ID	Price (USD)
`gpt-4o` (retired)	input $2.25 · output $9 (per 1M tokens)
`gpt-4o-2024-05-13` (retired)	input $4.5 · output $13.5 (per 1M tokens)

Model launch

3 models launched · pricingText

Model ID	Price (USD)
`gpt-3.5-turbo-0125` (retired)	input $0.45 · output $1.35 (per 1M tokens)
`text-embedding-3-large` (retired)	input $0.117 · output $0.117 (per 1M tokens)
`text-embedding-3-small` (retired)	input $0.018 · output $0.018 (per 1M tokens)

Model launch

3 models launched · pricingText

Model ID	Price (USD)
`gpt-3.5-turbo-1106` (retired)	input $0.9 · output $1.8 (per 1M tokens)

Audio

Model ID	Price (USD)
`tts-1` (retired 2026-09-03)	$13.5 per 1M characters of input text
`tts-1-hd` (retired 2026-09-03)	$27 per 1M characters of input text

Model launch

gpt-3.5-turbo-16k launched · pricing

Model ID	Price (USD)
`gpt-3.5-turbo-16k` (retired)	input $2.7 · output $8.1 (per 1M tokens)

Model launch

3 models launched · pricingText

Model ID	Price (USD)
`gpt-3.5-turbo` (retired)	input $0.45 · output $1.35 (per 1M tokens)
`text-embedding-ada-002` (retired)	input $0.09 · output $0.09 (per 1M tokens)

Audio

Model ID	Price (USD)
`whisper-1`	$0.0271 per minute of audio

Changelog

Contact