An LLM that costs very little during testing can produce a sizable monthly bill once a chatbot handles thousands of conversations or a coding agent processes large project files. Before committing to GPT, Claude, Gemini, DeepSeek, or another API, it helps to compare the cost of the specific models under consideration against the requests your application will actually make.
The quick comparison below lists current input, cached-input, and output prices for major LLMs alongside their context windows. The provider tables extend the reference to additional models and API pricing for image, audio, and video generation.
Last Updated: Oct 25, 2026
Quick LLM API Pricing Comparison
Standard direct-provider rates in USD per million text tokens.
| Model (Context) | Input | Cached Input | Output |
|---|---|---|---|
| GPT-6 Astra 1.05M context | $10.00 >272K: $20 | $1.00 >272K: $2 | $50.00 >272K: $75 |
| GPT-6.1 Sol 1.05M context | $2.00 >272K: $4 | $0.10 >272K: $0.20 | $10.00 >272K: $15 |
| GPT-6 Luna 1.05M context | $0.10 >272K: $0.20 | $0.01 >272K: $0.02 | $0.50 >272K: $0.75 |
| Claude Fable 5.1 1M context | $10.00 | $0.25 | $50.00 |
| Claude Opus 5.5 1M context | $4.00 | $0.20 | $20.00 |
| Claude Sonnet 5.5 1M context | $2.00 | $0.10 | $10.00 |
| Claude Haiku 5.5 1M context | $0.10 >100K: $0.50 | $0.01 >100K: $0.05 | $0.50 >100K: $2.50 |
| Gemini 4 Argon 1M context | $2.00 | $0.10 | $10.00 |
| Gemini 3.8 Flash 1M context | $0.75 | $0.075 | $3.75 |
| Gemini 3.1 Pro Preview 1M context | $2.00 >200K: $4 | $0.20 >200K: $0.40 | $12.00 >200K: $18 |
| Grok 4.7 500K context | $2.00 ≥200K: $4 | $0.50 ≥200K: $1 | $6.00 ≥200K: $12 |
| DeepSeek V4.1 Flash 1M context | $0.30 off-peak: $0.15 | $0.006 off-peak: $0.003 | $1.20 off-peak: $0.60 |
| Qwen3.8-Max 1M context | $2.00 | $0.25 | $6.00 |
| Qwen3.8-Flash 1M context | $0.15 | $0.016 | $0.47 |
| GLM-5.3 1M context | $1.40 | $0.26 | $4.40 |
| GLM-5.3-Flash 1M context | $0.15 | $0.03 | $0.50 |
| MiniMax M3 1M context | $0.30 >512K: $0.60 | $0.06 >512K: $0.12 | $1.20 >512K: $2.40 |
| MiniMax M3.1 Flash 1M context | $0.15 | $0.015 | $0.60 |
| Kimi K3.1 Pro 1M context | $2.00 | $0.20 | $10.00 |
| Mistral Large 4 1M context | $1.36 | $0.14 | $4.18 |
API Pricing by Provider
GPT-6 Astra API Pricing
Cache Write shows the 30-minute rate. A 12-hour write costs 2.5 times the regular input rate.
Above 272K input tokens, the higher long-context tier applies to the entire request.
| Processing | Input | Cached Input | Cache Write | Output |
|---|---|---|---|---|
| Standard, up to 272K input | $10.00 | $1.00 | $12.50 | $50.00 |
| Standard, over 272K input | $20.00 | $2.00 | $25.00 | $75.00 |
| Batch / Flex, up to 272K input | $5.00 | $0.50 | $6.25 | $25.00 |
| Batch / Flex, over 272K input | $10.00 | $1.00 | $12.50 | $37.50 |
| Fast, up to 272K input | $20.00 | $2.00 | $25.00 | $100.00 |
| Fast, over 272K input | $40.00 | $4.00 | $50.00 | $150.00 |
| Ultrafast, up to 272K input | $60.00 | $6.00 | $75.00 | $300.00 |
| Ultrafast, over 272K input | $120.00 | $12.00 | $150.00 | $450.00 |
OpenAI GPT Models API Pricing
GPT-6.1 also has Fast (2× Standard), Ultrafast (6× Standard), and Batch/Flex (0.5× Standard) processing prices.
| Model (Context) | Input | Cached Input | Output |
|---|---|---|---|
| GPT-6.1 Sol ≤272K input 1.05M context | $2.00 | $0.10 | $10.00 |
| GPT-6.1 Sol >272K input 1.05M context | $4.00 | $0.20 | $15.00 |
| GPT-6.1 Terra ≤272K input 1.05M context | $1.00 | $0.05 | $5.00 |
| GPT-6.1 Terra >272K input 1.05M context | $2.00 | $0.10 | $7.50 |
| GPT-6 Sol ≤272K input 1.05M context | $2.00 | $0.20 | $10.00 |
| GPT-6 Sol >272K input 1.05M context | $4.00 | $0.40 | $15.00 |
| GPT-6 Luna ≤272K input 1.05M context | $0.10 | $0.01 | $0.50 |
| GPT-6 Luna >272K input 1.05M context | $0.20 | $0.02 | $0.75 |
| GPT-5.6 Sol ≤272K input 1.05M context | $4.00 | $0.40 | $20.00 |
| GPT-5.6 Sol >272K input 1.05M context | $8.00 | $0.80 | $30.00 |
| GPT-5.6 Terra ≤272K input 1.05M context | $2.00 | $0.20 | $12.00 |
| GPT-5.6 Terra >272K input 1.05M context | $4.00 | $0.40 | $18.00 |
| GPT-5.6 Luna ≤272K input 1.05M context | $0.20 | $0.02 | $1.20 |
| GPT-5.6 Luna >272K input 1.05M context | $0.40 | $0.04 | $1.80 |
| GPT-5.6 Cyber 400K context | $12.50 | $1.25 | $75.00 |
| gpt-5.5 ≤272K input 1.05M context | $5.00 | $0.50 | $30.00 |
| gpt-5.5 >272K input 1.05M context | $10.00 | $1.00 | $45.00 |
| gpt-5.5-pro 1.05M context | $30.00 | – | $180.00 |
| gpt-5.4 ≤272K input 1.05M context | $2.50 | $0.25 | $15.00 |
| gpt-5.4 >272K input 1.05M context | $5.00 | $0.50 | $22.50 |
| gpt-5.4-pro ≤272K input 1.05M context | $30.00 | – | $180.00 |
| gpt-5.4-pro >272K input 1.05M context | $60.00 | – | $270.00 |
| gpt-5.4-mini 400K context | $0.75 | $0.075 | $4.50 |
| gpt-5.4-nano 400K context | $0.20 | $0.02 | $1.25 |
| gpt-5.2 400K context | $1.75 | $0.175 | $14.00 |
| gpt-5.2-pro 400K context | $21 | – | $168 |
| gpt-5.1 400K context | $1.25 | $0.125 | $10.00 |
| gpt-5 400K context | $1.25 | $0.125 | $10.00 |
| gpt-5-mini 400K context | $0.25 | $0.025 | $2.00 |
| gpt-5-nano 400K context | $0.05 | $0.005 | $0.40 |
| gpt-5-pro 400K context | $15.00 | – | $120.00 |
| gpt-5.3-chat-latest 400K context | $1.75 | $0.175 | $14.00 |
| gpt-5.3-codex | $1.75 | $0.175 | $14.00 |
| gpt-5.1-codex-max | $1.25 | $0.125 | $10.00 |
| gpt-5.1-codex-mini | $0.25 | $0.025 | $2.00 |
| codex-mini-latest | $1.50 | $0.375 | $6.00 |
| gpt-5-search-api | $1.25 | – | $10.00 |
| gpt-4.1 1M context | $2.00 | $0.50 | $8.00 |
| gpt-4.1-mini 1M context | $0.40 | $0.10 | $1.60 |
| gpt-4.1-nano 1M context | $0.10 | $0.025 | $0.40 |
| gpt-4o 128K context | $2.50 | $1.25 | $10.00 |
| gpt-4o-mini 128K context | $0.15 | $0.075 | $0.60 |
OpenAI Reasoning Models API Pricing (Legacy)
| Model | Context Window | Input/1M Tokens | Output/1M Tokens |
|---|---|---|---|
| o4-mini | 200K | $1.10 | $4.40 |
| o4-mini-deep-research | 200K | $2.00 | $8.00 |
| o3-pro | 200K | $20.00 | $80.00 |
| o3 | 200K | $2.00 | $8.00 |
| o3-deep-research | 200K | $10.00 | $40.00 |
| o3-mini | 200K | $1.10 | $4.40 |
| o1-pro | 200K | $150.00 | $600.00 |
| o1 | 200K | $15.00 | $60.00 |
| o1-mini | 128K | $1.10 | $4.40 |
OpenAI Image Generation API Pricing
| Model | Input/1M Tokens | Output/1M Tokens |
|---|---|---|
| GPT-image-2.5-sunburst | $5.00 (text) $8.00 (image) | $30.00 |
| GPT-image-2.5-flare | $5.00 (text) $8.00 (image) | $30.00 |
| GPT-image-2 | $5.00 (text) $8.00 (image) | $30.00 |
| GPT-image-1.5 | $5.00 | $10.00 |
| GPT-image-1 | $5.00 | N/A |
| GPT-image-1-mini | $2.00 | N/A |
OpenAI Voice & Audio API Pricing
| Model | Input/1M Tokens | Output/1M Tokens |
|---|---|---|
| gpt-live-1 | N/A | $0.05 / minute |
| gpt-realtime-2.1 | $4.00 (text) $5.00 (image) $32.00 (audio) | $24.00 (text) $64.00 (audio) |
| gpt-realtime-2.1-mini | $0.60 (text) $0.80 (image) $10.00 (audio) | $2.40 (text) $20.00 (audio) |
| gpt-realtime-translate | N/A | $0.034 / minute |
| gpt-transcribe | N/A | $0.0045 / minute |
| gpt-live-transcribe | N/A | $0.017 / minute |
| gpt-realtime-whisper | N/A | $0.017 / minute |
Claude API Pricing
Claude Haiku 5.5 uses its higher pricing tier above 100K prompt tokens, including cached tokens. Cache writes cost 1.25× the regular input rate for five-minute retention or 2× for one hour.
| Model (Context) | Input | Cached Input | Output |
|---|---|---|---|
| Claude Fable 5.1 1M context | $10.00 | $0.25 | $50.00 |
| Claude Mythos 5.1 (limited availability) 1M context | $10.00 | $0.25 | $50.00 |
| Claude Opus 5.5 1M context | $4.00 | $0.20 | $20.00 |
| Claude Opus 5 1M context | $5.00 | $0.50 | $25.00 |
| Claude Opus 4.8 1M context | $5.00 | $0.50 | $25.00 |
| Claude Sonnet 5.5 1M context | $2.00 | $0.10 | $10.00 |
| Claude Haiku 5.5 ≤100K input 1M context | $0.10 | $0.01 | $0.50 |
| Claude Haiku 5.5 >100K input 1M context | $0.50 | $0.05 | $2.50 |
Gemini API Pricing
Gemini 4 Argon changes to $4 input / $0.20 cached input / $20 output on December 1, 2026. Gemini 3.8 Flash also has higher regular rates scheduled for January 1, 2027. Gemini 4 Argon uses implicit caching only.
| Model (Context) | Input | Cached Input | Output |
|---|---|---|---|
| Gemini 4 Argon 1M context | $2.00 through 11/30/2026 $4.00 starting 12/01/2026 | $0.10 $0.20 from Dec 1, 2026 | $10.00 through 11/30/2026 $20.00 starting 12/01/2026 |
| Gemini 3.8 Flash 1M context | $0.75 | $0.075 | $3.75 |
| Gemini 3.6 Flash 1M context | $0.75 | $0.075 | $3.75 |
| Gemini 3.5 Flash-Lite 1M context | $0.30 | $0.03 | $2.50 |
| Gemini 3.5 Transcribe Live | $3.50 or $0.005/min | – | $21.00 or $0.004/min |
| Gemini 3.5 Transcribe | $2.00 or $0.003/min | – | $12.00 or $0.002/min |
| Gemini 3.1 Pro Preview ≤200K input 1M context | $2.00 | $0.20 | $12.00 |
| Gemini 3.1 Pro Preview >200K input 1M context | $4.00 | $0.40 | $18.00 |
| Gemini 3.1 Flash-Lite 1M context | $0.25 (text/image/video) $0.50 (audio) | $0.025 (text/image/video) $0.05 (audio) | $1.50 |
| Gemini 3.1 Flash Live | $0.75 (text) $3.00 or $0.005/min (audio) $1.00 or $0.002/min (image/video) | – | $4.50 (text) $12.00 or $0.018/min (audio) |
| Gemini 3 Flash Preview 1M context | $0.50 (text / image / video) $1.00 (audio) | $0.05 | $3.00 |
| Gemini 2.5 Pro ≤200K input 1M context | $1.25 | $0.125 | $10.00 |
| Gemini 2.5 Pro >200K input 1M context | $2.50 | $0.25 | $15.00 |
| Gemini 2.5 Flash 1M context | $0.30 (text/image/video) $1.00 (audio) | $0.03 (text/image/video) $0.10 (audio) | $2.50 |
| Gemini 2.5 Flash-Lite 1M context | $0.10 (text/image/video) $0.50 (audio) | $0.01 (text/image/video) $0.03 (audio) | $0.40 |
Nano Banana Pricing
| Model | Free Tier | Input/1M Tokens | Output/1M Tokens |
|---|---|---|---|
| Nano Banana 2.1 | Not available | $1.50 | $7.50 (text and thinking) $30.00 (images) $0.0336 per 1K image $0.0504 per 2K image $0.0756 per 4K image |
| Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image) | Not available | $0.25 | $1.50 (text and thinking) $30.00 (images) $0.034 per 1K image |
| Nano Banana 2 (Gemini 3.1 Flash Image) | Not available | $0.50 | $3.00 (text and thinking) $60.00 (images) $0.067 per 1K image $0.101 per 2K image $0.151 per 4K image |
| Nano Banana 2 Pro (Gemini 3 Pro Image) | Not available | $2.00 (text) $0.0011 (image) | $12.00 (text and thinking) $120.00 (images) $0.134 per 1K/2K image $0.24 per 4K image |
Gemini Omni Pricing
| Model | Free Tier | Input/1M Tokens | Output/1M Tokens |
|---|---|---|---|
| Gemini Omni 1.1 Flash | Not available | $1.50 (text/image/video/audio) | $9.00 (text) $17.50 (video) |
Gemini Live
| Model | Free Tier | Input | Output |
|---|---|---|---|
| Gemini 3.8 Live Gemini 3.8 Live Extended Thinking | Free of charge | $0.75 (text) $3.00 or $0.005/min (audio) $1.00 or $0.002/min (image/video) | $4.50 (text) $12.00 or $0.018/min (audio) |
| Gemini 3.5 Live Translate | Free of charge | $3.50 or $0.0053/min | $21.00 or $0.0315/min |
Gemini 2.5 Flash Native Audio API Pricing
| Model | Free Tier | Input/1M Tokens | Output/1M Tokens |
|---|---|---|---|
| Gemini 2.5 Flash Native Audio | Not available | $0.50 (text) $3.00 (audio/video) | $2.00 (text) $12.00 (audio) |
Gemini TTS Pricing
| Model | Free Tier | Input/1M Tokens | Output/1M Tokens |
|---|---|---|---|
| Gemini 3.1 Flash TTS Preview | Free of charge | $1.00 (text) | $20.00 (audio) |
| Gemini 2.5 Flash Preview TTS | Free of charge | $0.50 (text) | $10.00 (audio) |
| Gemini 2.5 Pro Preview TTS | Not available | $1.00 (text) | $20.00 (audio) |
Google Imagen 4 & 3 API Pricing
| Model | Paid Tier, per Image in USD |
|---|---|
| Imagen 4 Fast | $0.02 |
| Imagen 4 Standard | $0.04 |
| Imagen 4 Ultra | $0.06 |
| Imagen 3 | $0.03 |
Gemini Embedding API Pricing
| Model | Paid Tier, per 1M tokens in USD |
|---|---|
| gemini-embedding-2 | $0.20 (Text) $0.45 (Image $0.00012 per image) $6.50 (Audio $0.00016 per second) $12.00 (Video $0.00079 per frame) |
| gemini-embedding-001 | $0.15 |
Google Veo 3.1 API Pricing
| Model | Paid Tier, per second in USD |
|---|---|
| Veo 3.1 Standard | $0.40 (720p and 1080p) $0.60 (4k) |
| Veo 3.1 Fast | $0.15 (720p and 1080p) $0.35 (4k) |
| Veo 3.1 Light | $0.05 (720p) $0.08 (1080p) |
Grok API Pricing
For Grok 4.7, requests at or above 200K input tokens are charged at the long-context rate for the entire request.
| Model (Context) | Input | Cached Input | Output |
|---|---|---|---|
| Grok 4.7 <200K input 500K context | $2.00 | $0.50 | $6.00 |
| Grok 4.7 ≥200K input 500K context | $4.00 | $1.00 | $12.00 |
| Grok Build 0.1 <200K input 256K context | $1.00 | $0.20 | $2.00 |
| Grok Build 0.1 ≥200K input 256K context | $2.00 | $0.40 | $4.00 |
| Grok 4.3 1M context | $1.25 | $0.20 | $2.50 |
| Grok 4.20 Multi-agent 1M context | $1.25 | $0.20 | $2.50 |
| Grok 4.20 Reasoning 1M context | $1.25 | $0.20 | $2.50 |
| Grok 4.20 1M context | $1.25 | $0.20 | $2.50 |
Grok Image, Video, and Voice API Pricing
| Model | Input / Processing | Output / Unit |
|---|---|---|
| Imagine Image 2.0 | $0.01 / image | $0.04/image (1K Low) $0.05/image (1.5K Low) $0.06/image (2K Low) $0.06/image (1K Medium) $0.07/image (1.5K Medium) $0.08/image (2K Medium) |
| Grok Imagine Image Quality | $0.01 / image | $0.05/image (1K) $0.06/image (1.5K) $0.07/image (2K) |
| Grok Imagine Image | $0.002 / image | $0.02/image (1K) $0.02/image (2K) |
| Grok Imagine Video 1.5 | $0.01 / image | $0.08/sec (480p) $0.14/sec (720p) $0.25/sec (1080p) |
| Grok Imagine Video 1.5 Lite | $0.01 / image | $0.02/sec (480p) $0.03/sec (720p) $0.14/sec (1080p) |
| Grok Imagine Video | $0.01/second $0.002/image | $0.05/sec (480p) $0.07/sec (720p) |
| Grok Voice 3.0 | N/A | $0.08/min ($4.80/hr) audio $0.004/text input |
| Text to Speech | N/A | $15.00 / 1M chars |
| Speech to Text | N/A | $0.10 / hr (REST) $0.20 / hr (Streaming) |
DeepSeek API Pricing
Peak hours: 01:00–04:00 and 06:00–10:00 UTC on weekdays excluding Chinese public holidays. Other hours use off-peak rates.
| Model (Context) | Input | Cached Input | Output |
|---|---|---|---|
| DeepSeek-V4.1-Flash-1017 1M context | $0.15 (OFF-PEAK) $0.30 (PEAK) | $0.003 off-peak $0.006 peak | $0.60 (OFF-PEAK) $1.20 (PEAK) |
| DeepSeek-V4-Pro-0813 1M context | $0.66 (OFF-PEAK) $1.32 (PEAK) | $0.022 off-peak $0.044 peak | $1.98 (OFF-PEAK) $3.96 (PEAK) |
Qwen API Pricing
International deployment prices. For Qwen3.8-Max, explicit cache reads cost $0.17/1M tokens versus $0.25/1M for implicit cache hits.
| Model (Context) | Input | Cached Input | Output |
|---|---|---|---|
| Qwen3.8-Max 1M context | $2.00 | $0.25 | $6.00 |
| Qwen3.8-Flash 1M context | $0.15 | $0.016 | $0.47 |
| Qwen3.7-Plus ≤256K input 1M context | $0.40 | $0.08 | $1.60 |
| Qwen3.7-Plus >256K input 1M context | $1.20 | $0.24 | $4.80 |
| Qwen3.7-Flash ≤32K input 1M context | $0.03 | $0.006 | $0.13 |
| Qwen3.7-Flash 32K–256K input 1M context | $0.10 | $0.02 | $0.40 |
| Qwen3.7-Flash >256K input 1M context | $0.20 | $0.04 | $0.80 |
Qwen Image, Video, and Omni API Pricing
Image prices use Singapore International rates. Qwen Image 3.0 Pro bills input images and generated images separately.
| Model | Input / Processing | Output / Unit |
|---|---|---|
| Qwen3.8-Omni-Flash | $0.15 | $0.47 |
| Qwen3.5-Omni-Plus | $11.00 (Audio) $1.40 (Text/Image/Video) | $44.00 (Text & Audio) $8.30 (Text) |
| Qwen Image 3.0 Pro | $0.003 / input image | $0.04 / output image (1K) $0.075 / output image (2K) |
| Qwen Image 2.1 Pro | N/A | $0.04/IMG |
| Qwen Image 2.1-Turbo | N/A | $0.016/IMG |
| Wan 3.0 Video | N/A | $0.05/sec (480p) $0.10/sec (720p) $0.20/sec (1080p) |
| Wan2.7-Image-Pro | N/A | $0.075/IMG Generation |
| HappyHorse-1.1-T2V | N/A | $0.07/sec (480p) $0.14/sec (720p) $0.24/sec (1080p) |
Mistral API Pricing
| Model (Context) | Input | Cached Input | Output |
|---|---|---|---|
| Mistral Large 4 1M context | $1.36 | $0.14 | $4.18 |
| Mistral Medium 3.5 256K context | $1.50 | $0.15 | $7.50 |
| Mistral Large 3 256K context | $0.50 | $0.05 | $1.50 |
| Ministral 3 – 3B 256K context | $0.10 | $0.01 | $0.10 |
| Ministral 3 – 8B 256K context | $0.15 | $0.015 | $0.15 |
| Ministral 3 – 14B 256K context | $0.20 | $0.02 | $0.20 |
| Mistral Small 4 128K context | $0.15 | $0.015 | $0.60 |
| Codestral 32K context | $0.30 | $0.03 | $0.90 |
| Codestral Embed 128K context | $0.15 | $0.015 | N/A |
Mistral OCR, Speech, and Embedding API Pricing
| Model | Input / Processing | Output / Unit |
|---|---|---|
| OCR 4.1 | $4 /1000 Pages | N/A |
| Voxtral Mini Transcribe 3 | $0.0035 /Min | N/A |
| Voxtral Mini Transcribe Realtime 3 | $0.008 /Min | |
| Voxtral Small Transcribe 3 | $0.004 /Min | N/A |
| Voxtral TTS 3 | $13 /M Chars | N/A |
| Voxtral Mini Transcribe 2 | $0.003 /Min | N/A |
| Voxtral TTS | $16 /M Chars | N/A |
Black Forest Labs FLUX API Pricing
FLUX Image API Pricing
| Model | Resolution / Task | API Price |
|---|---|---|
| FLUX 3 Image 1 | 768 × 768 | $0.041 / image |
| FLUX 3 Image 1 | 1K | $0.048 / image |
| FLUX 3 Image 1 | 1.5K | $0.07 / image |
| FLUX 3 Image 1 | 2K | $0.10 / image |
| FLUX 3 Image 1 | 4K | $0.607 / image |
| FLUX 3 Image Klein | 1K | from $0.009 / image |
| FLUX.2 [pro] | Text to image | from $0.03 / MP |
| FLUX.2 [pro] | Image editing | from $0.045 / MP |
| FLUX.2 [max] | Generation / editing | from $0.07 / MP |
| FLUX.2 [flex] | Text to image | $0.05 / MP |
| FLUX.2 [flex] | Image editing | $0.10 / MP |
| FLUX.2 [klein] 4B | Text to image | from $0.014 / image |
| FLUX.2 [klein] 9B | Text to image | from $0.015 / image |
| FLUX.1 Kontext [pro] | Generation / editing | $0.04 / image |
FLUX 3 Video API Pricing
Rates in USD per generated second. Draft is an HD preview.
| Mode | HD | FHD | Draft |
|---|---|---|---|
| Text to Video | $0.17 | $0.29 | $0.06 |
| Image to Video | $0.17 | $0.29 | $0.06 |
| Video Continuation | $0.43 | $0.54 | $0.12 |
Meta AI Muse API Pricing
| Model | Context Window | Input/1M Tokens | Output/1M Tokens |
|---|---|---|---|
| Muse Wonder | 1M | $4.00 | $20.00 |
| Muse Spark 1.3 (Contributor) | 1M | $0.10 | $0.20 |
| Muse Spark 1.3 | 1M | $1.25 | $4.25 |
| Muse Image | N/A | N/A | $0.01/image |
| Muse Voice Transcribe | N/A | N/A | $3.00/1,000 min $0.18/hour |
GLM API Pricing
| Model (Context) | Input | Cached Input | Output |
|---|---|---|---|
| GLM-5.3 1M context | $1.40 | $0.26 | $4.40 |
| GLM-5.3-Flash 1M context | $0.15 | $0.03 | $0.50 |
| GLM-5.2 1M context | $1.40 | $0.26 | $4.40 |
| GLM-5.1 200K context | $1.40 | $0.26 | $4.40 |
| GLM-5 200K context | $1.00 | $0.20 | $3.20 |
| GLM-5-Turbo 200K context | $1.20 | – | $4.00 |
| GLM-5-Code 200K context | $1.20 | – | $5.00 |
| GLM-5V-Turbo 200K context | $1.20 | – | $4.00 |
| GLM-OCR | $0.03 | – | $0.03 |
| GLM-4.7 128K context | $0.60 | $0.11 | $2.20 |
| GLM-4.7-FlashX 128K context | $0.07 | $0.01 | $0.40 |
| GLM-4.5 128K context | $0.60 | $0.11 | $2.20 |
Kimi K3.1 & K3 API Pricing
K3-series cache writes have separate five-minute and one-hour retention rates.
| Model (Context) | Input | Cached Input | Output |
|---|---|---|---|
| kimi-k3.1 Pro 1M context | $2.00 | $0.20 | $10.00 |
| kimi-k3.1 Pro-highspeed 1M context | $4.00 | $0.20 | $20.00 |
| kimi-k3 1M context | $3.00 | $0.30 | $15.00 |
| kimi-k2.7-code-highspeed 262,144 context | $1.90 | $0.19 | $8.00 |
| kimi-k2.7-code 262,144 context | $0.95 | $0.19 | $4.00 |
| kimi-k2.6 262,144 context | $0.95 | $0.16 | $4.00 |
| kimi-k2.5 262,144 context | $0.60 | – | $3.00 |
| kimi-k2-thinking 262,144 context | $0.60 | – | $2.50 |
MiniMax M3 & M2.7 API Pricing
| Model (Context) | Input | Cached Input | Output |
|---|---|---|---|
| MiniMax-M3.1-Flash 1M context | $0.15 | $0.015 | $0.60 |
| Minimax M3 ≤512K input 1M context | $0.30 | $0.06 | $1.20 |
| Minimax M3 >512K input 1M context | $0.60 | $0.12 | $2.40 |
| Minimax M2.7 204,800 context | $0.30 | $0.06 | $1.20 |
| Minimax M2.7 Highspeed 204,800 context | $0.60 | $0.06 | $2.40 |
Minimax Video/Audio/Image Generation API Pricing
| Model | Unit Price |
|---|---|
| MiniMax H3 | $0.13/s, 2K |
| MiniMax H3 | $0.09/s, 768P |
| MiniMax H3 Max | $0.05/s, 480P |
| MiniMax H3 Max | $0.08/s, 768P |
| Speech Recognition | $0.38/hour |
| speech-2.8-hd | $100/M characters |
| speech-2.8-turbo | $60/M characters |
| image-01 | $0.0035 per image |
Perplexity API Pricing
| Model | Context Window | Input/1M Tokens | Output/1M Tokens |
|---|---|---|---|
| Sonar | 128K | $1.00 | $1.00 |
| Sonar Pro | 200K | $3.00 | $15.00 |
| Sonar Reasoning Pro | 200K | $2.00 | $8.00 |
| Sonar Deep Research | 128K | $2.00 | $8.00 |
Cohere API Pricing
| Model | Context Window | Input/1M Tokens | Output/1M Tokens |
|---|---|---|---|
| Command A | 256K | $2.50 | $10.00 |
| Command R | 128K IN/4K OUT | $0.15 | $0.60 |
| Command R7B | 128K | $0.0375 | $0.15 |
| Transcribe | N/A | N/A | $3.75/hour/instance |
| Embed 5 Fast | 128K | N/A | $0.08 |
| Embed 5 Pro | 128K | N/A | $0.12 |
| Rerank 5 Fast | N/A | N/A | $0.02 |
| Rerank 5 Pro | N/A | N/A | $0.05 |
| Parse 5 | N/A | N/A | $1.50/1K pages |
How to Compare LLM API Pricing
For a workload estimate, multiply uncached input and generated output tokens by their respective rates. Include the model’s cached-input price for repeated context only when a request actually produces a cache hit.
Reasoning output tokens, search or tool calls, and provider-specific long-context tiers can change the final bill. Estimate a typical production request or monthly token mix, not only the price of a single input token.
Text API cost = (uncached input × input rate + cached input × cache-read rate + output × output rate) ÷ 1,000,000
Cached Input and Cache-Write Pricing
Cached Input applies to prompt tokens actually served from a cache. A repeated prompt that misses the cache is charged at the regular input rate. Cache-write and storage pricing varies by provider.
OpenAI and Anthropic can charge a cache-write token rate for creating a reusable prefix. A write is billed at its applicable input-token rate, not as an additional surcharge on those same tokens. Gemini explicit caches can incur token-storage charges by duration; DeepSeek prices vary with peak and off-peak hours.
Example: A GPT-6.1 Sol request with 8,000 cached input tokens, 2,000 uncached input tokens, and 2,000 output tokens costs $0.0248 at the short-context Standard rate: ($0.10 × 8,000 + $2 × 2,000 + $10 × 2,000) ÷ 1,000,000. This assumes the cached prefix already exists; creating it earlier incurs its own cache-write charge.
LLM API Pricing FAQ
Does prompt caching always reduce API costs?
No. A cache hit can lower the cost of reused input, but the first write can cost more than a normal input request. Savings depend on how often the same prompt prefix is reused before the cache expires.
Are ChatGPT, Claude, and Gemini subscriptions the same as API pricing?
No. Consumer subscriptions and API usage use different billing systems.
A ChatGPT, Claude, or Gemini subscription pays for access to the company’s consumer application and the features included with that plan. API access is generally billed through a developer platform according to token usage or other API-specific units.
Why can long-context API requests cost more?
Long prompts consume more input tokens, so they cost more even when the per-token rate stays unchanged. Some models also apply a higher pricing tier after the request crosses a specified context threshold.
For example, GPT-6 Astra and GPT-5.6 apply higher rates when input exceeds 272K tokens. Gemini 3.1 Pro Preview uses higher rates above 200K input tokens.
Are image and video API prices charged per token?
Not always. Depending on the API, image generation can be billed per image or megapixel, and video generation per second. Some multimodal models also use input and output token rates.
See Also:
- What Are The Rate Limits For OpenAI API?
- Compare AI Costs: Free LLM API Price Calculator
- ChatGPT Token Limit: Free, Plus, Pro, and OpenAI API Limits
- AI Coding Plan Comparison: Prices, Limits, and Features
- LLM Context Window Comparison: GPT, Claude, Gemini & More
- Gemini API Free Tier Limits: Requests Per Day & Setup
Changelog:
10/26/2026
- Added GPT-6.1 Terra
10/25/2026
- Sora 2 and Sora 2 Pro APIs shut down on September 24. Gemini 2.0 Flash was discontinued, and Gemini 3.5/3.7 Flash requests now use replacement models.
10/25/2026
- Added Black Forest Labs FLUX 3 Image, FLUX.2, and FLUX 3 Video API prices.
10/23/2026
- Added Kimi 3.1 Pro
10/21/2026
- Added Rerank 5
10/20/2026
- Added Voxtral 3
10/19/2026
- Added MiniMax-M3.1-Flash
10/17/2026
- Added DeepSeek-V4.1-Flash-1017
10/14/2026
- Added Grok Voice 3.0
10/13/2026
- Added Gemini Argon
10/12/2026
- Added Muse Wonder
10/09/2026
- Added Qwen Image 2.1 Pro
10/08/2026
- Added GPT-6.1 Sol Ultrafast mode.
- Added Grok Imagine Video 1.5 Lite.
10/07/2026
- Added Claude Haiku 5.5
10/06/2026
- Added Mistral Large 4
- Added Nano Banana 2.1
09/29/2026
- Added GPT-6.1 Sol
09/28/2026
- Added Sonnet 5.5
09/22/2026
- Added Grok 4.7
- Added Opus 5.5
- Added GPT‑6 Sol and Luna
09/18/2026
- Updated Grok models
09/17/2026
- Added Qwen3.8-Omni-Flash
09/15/2026
- Added Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking
09/10/2026
- Added DeepSeek-V4.1-Flash
- Added gpt-live-1
09/08/2026
- Corrected context-window and long-context pricing threshold labels for GPT and Gemini models.
- Updated MiniMax M3, Cohere Command A, and Devstral 2 API pricing.
- Added GPT-Image-2.5 Flare and Sunburst.
09/03/2026
- Added GPT-6 Astra
09/02/2026
- Added Muse Voice Transcribe
- Added Claude Fable 5.1 and Claude Mythos 5.1
- Added Muse Spark 1.3
08/27/2026
- Added Gemini Omni 1.1 Flash
- Added cohere models
08/26/2026
- Added GLM-5.3-Flash and Qwen3.8-Flash
- Added Muse Image and Gemini 3.5 Transcribe
08/21/2026
- Added DeepSeek-v4-flash-vision-exp.
- Updated GPT-5.6 Sol.
08/19/2026
- Added GLM 5.3.
08/16/2026
- Updated DeepSeek API pricing.
08/13/2026
- Added Mistral OCR 4.1
- Added Gemini 3.7 Flash
- Added GLM 5.3
- Added Minimax Music-3.0
08/12/2026
- Added DeepSeek-V4-Pro-0813
- Added Grok 4.6
08/11/2026
- Added Grok Imagine Image 2.0
08/06/2026
- Added Wan 3.0 Video
08/05/2026
- Added Muse Spark 1.2
08/02/2026
- Added Qwen3.8-Max
07/31/2026
- Added Minimax H3.
07/30/2026
- Updated GPT 5.6 pricing.
07/29/2026
- Added Grok Voice Think Fast 2.0
07/28/2026
- Cleanup
- Added GPT-Live-Transcribe and GPT-Transcribe
07/24/2026
- Added Claude Opus 5
07/21/2026
- Added Gemini 3.6 Flash and 3.5 Flash-Lite
07/16/2026
- Added Kimi K3.
07/09/2026
- Updated for GPT-5.6.
07/08/2026
- Added Grok 4.5.
07/06/2026
- Added gpt-realtime-2.1-mini
07/01/2026
- Added kimi-k2.7-code-highspeed
06/30/2026
- Added Claude Sonnet 5
- Added Nano Banana Lite
06/26/2026
- Added GPT 5.6
06/16/2026
- Added OCR 4
06/16/2026
- Added GLM-5.2
06/13/2026
- Updated Minimax M3
06/12/2026
- Added kimi-k2.7-code
06/09/2026
- Added Claude Fable/Mythos 5.
- Added Gemini 3.5 Live Translate.
06/07/2026
- Updated Deepseek API.
06/04/2026
- Added grok-imagine-video-1.5-preview
06/01/2026
- Added MiniMax M3
- Added Qwen3.7-Plus and HappyHorse-1.0-T2V
05/28/2026
- Added Claude Opus 4.8
05/21/2026
- Added Qwen3.7-Max
05/19/2026
- Added Gemini 3.5 Flash.
05/13/2026
- Added Claude Opus 4.7 Fast Mode
05/08/2026
- Added grok-imagine-image-quality
05/07/2026
- Added OpenAI audio models
05/01/2026
- Added Grok 4.3
04/29/2026
- Updated for Mistral Medium 3.5
04/27/2026
- Updated for DeepSeek v4
04/25/2026
- Updated
04/24/2026
- Added Deeseek V4 models
04/23/2026
- Added GPT-5.5
04/22/2026
- Added Gemini Embedding 2
04/22/2026
- Added GPT-image-2
04/21/2026
- Added Qwen3.6-Max-Preview
04/20/2026
- Added Kimi K2.6
04/18/2026
- Cleanup Grok APIs.
04/16/2026
- Added Claude Opus 4.7
04/15/2026
- Added Gemini 3.1 Flash TTS Preview
04/08/2026
- Added GLM-5.1
04/01/2026
- Added Perplexity Sonar models.
- GLM-5V-Turbo.
03/31/2026
- Added Veo 3.1 Light
03/26/2026
- Added Voxtral TTS and Gemini 3.1 Flash Live
03/23/2026
- Updated for Grok 4.2-0309
03/18/2026
- Added Minimax M2.7.
03/17/2026
- Added GPT-5.4 mini and nano
03/16/2026
- Added Mistral Small 4.0
03/16/2026
- Added GLM-5-Turbo
03/05/2026
- Added GPT-5.4
03/03/2026
- Added Gemini 3.1 Flash-Lite
- Added GPT-5.3 Instant
02/26/2026
- Added Nano Banana 2 (Gemini 3.1 Flash Image Preview)
02/19/2026
- Added Gemini Pro 3.1
02/18/2026
- Added Sonnet 4.6
02/11/2026
- Added GLM-5
02/06/2026
- Added Opus 4.6 & Voxtral Mini Transcribe 2
01/29/2026
- Added grok-imagine-video and Kimi 2.5.
12/23/2025
- Updated GLM-4.7
12/17/2025
- Added Gemini 3 Flash
12/16/2025
- Added GPT-image-1.5
12/11/2025
- Added GPT 5.2
12/09/2025
- Added Devstral 2
12/02/2025
- Added Mistral Large 3
12/01/2025
- Updated for DeepSeek-V3.2-Speciale
11/24/2025
- Updated for Claude Opus 4.5
11/20/2025
- Added Nano Banana Pro.
- Added Grok 4.1 Fast.
- Removed legacy models.
11/19/2025
- Added more models
11/18/2025
- Added Gemini 3 Pro
11/14/2025
- Added GPT-5.1
- Updated Qwen models
- Updated Kimi models
- Added Minimax M2
11/08/2025
- Added codex-mini
10/15/2025
- Update for Claude Haiku 4.5 & Veo 3.1.
10/07/2025
- Added Sora 2 and more OpenAI APIs.
09/29/2025
- Updated for Claude 4.5 and DeepSeek-V3.2
09/19/2025
- Updated price
08/21/2025
- Updated for DeepSeek-V3.1
08/20/2025
- Added GLM-4.5 and Kimi K2 models.
08/14/2025
- Added Imagen 4 Fast
08/07/2025
- Added GPT-5
08/05/2025
- Added Claude Opus 4.1
07/15/2025
- Added Voxtral model family
07/14/2025
- Added Gemini Embedding
07/11/2025
- Added Grok 4
06/26/2025
- Added o3-deep-research and o4-mini-deep-research
06/25/2025
- Added Imagen 4 Ultra and Imagen 4 Standard
06/20/2025
- Added Gemini 2.5 Flash-Lite
06/10/2025
- Added o3-pro
- Added Mistral Magistral models.
06/10/2025
- OpenAI dropped the price of o3 by 80%
05/22/2025
- Updated for Claude 4
05/21/2025
- Updated Mistral models
05/21/2025
- Added Gemini 2.5 Flash Native Audio
05/08/2025
- Added Mistral Medium 3
04/23/2025
- Added OpenAI Image Generation
04/18/2025
- Added Gemini 2.5 Flash
- Updated Cohere models
04/16/2025
- Added o4-mini
04/14/2025
- Added gpt-4.1 family
04/11/2025
- Updated DeepSeek
04/11/2025
- Added Google Imagen 3 and Veo 2.
04/10/2025
- Added Qwen models
- Added Grok 3
04/05/2025
- Added Gemini Pro 2.5
03/19/2025
- Added o1-pro
02/28/2025
- Added GPT-4.5
02/24/2025
- Added Claude 3.7
02/21/2025
- Added Grok
02/06/2025
- Added Gemini 2.0 Flash and Gemini 2.0 Flash-Lite
02/02/2025
- Added o3-mini
01/31/2025
- Added DeepSeek v3 and DeepSeek R1.
12/18/2024
- o1 in the API comes with support for function calling, developer messages, Structured Outputs, and vision capabilities.
12/07/2024
- Added Llama 3.3
11/05/2024
- Added Claude Haiku 3.5
11/03/2024
- Added Gemini 1.5 Flash-8B
10/04/2024
- Added Llama 3.2
- Added gpt-4o-realtime-preview
09/25/2024
- Updated Google Gemini
09/13/2024
- Added OpenAI’s latest model: o1.
08/07/2024
- Added gpt-4o-2024-08-06, the latest gpt-4o snapshot that supports Structured Outputs
07/25/2024
- Updated Mistral Large 2
07/24/2024
- Added Llama 3.1 405B
07/20/2024
- Added GPT-4o-mini
- Updated prices
07/13/2024
- Updated
- Added Cohere’s Command API
