
Major LLM API Pricing Comparison [Nov 2026]
Cached-input and cache-write rates (with retention and minimum lengths) are tabulated separately in Prompt Caching Pricing Compared. Batch API discounts (asynchronous batch) are compared in Batch API discounts compared. Speech (TTS and transcription) rates are compared separately in Speech, TTS and transcription API pricing. Frontier models are compared side by side in Frontier models compared.
We reference each provider's official pricing page and list only re-verified figures. All prices in USD / 1M tokens. Use English calculator to estimate costs. See GPT-5.6 Luna deep dive β, or the glossary entry on prompt caching (cache hits).
Re-fetched the official pricing pages on 2026-11-05 for Azureβs DeepSeek tab (only the V4 Pro / V4 Flash generations are listed β there is no DeepSeek-V4.1-Flash row), Grok (docs.x.ai), Qwen (Alibaba Cloud Model Studio), Meta Model API and StepFun: no price changes. OpenAI, Anthropic, Google, DeepSeek, Z.ai and Azure OpenAI were verified on 2026-11-03.
OpenAI
π Standard, short-context rate. Long-context: see official page
How to read the status column: Current = latest of its line and served via API today / Legacy = still served, but a successor exists (rows are never removed) / Retired = no longer served / Limited access = approval or invite only / Preview = preview stage. Lines are judged by model family (e.g. Sol / Terra / Luna, Gemini 3.x vs 4). Prices are as published at the time.
| Model | Input | Cached Input | Output | Notes | Status |
|---|---|---|---|---|---|
| GPT-6.2 Astra | $8.00 | $0.40 | $40.00 | Flagship (announced 2026-11-17). Long context $16.00/$0.80/$60.00. 20% cheaper in/out and 60% cheaper cached than GPT-6 Astra. Cache writes $10 (30m) / $20 (12h) β timeline | Current |
| GPT-6.2 Sol | $2.00 | $0.10 | $10.00 | General purpose (announced 2026-11-17). Same price as GPT-6.1 Sol. Long context $4.00/$0.20/$15.00 β timeline | Current |
| GPT-6.2 Astra for Law | $8.00 | $0.40 | $40.00 | Legal configuration (released 2026-11-17). Available to ChatGPT Enterprise customers through the Trusted Access Program (API included). Promotional rates through at least 2026-12-31. Cache writes $10 (30m) / $20 (12h). β timeline | Limited access |
| GPT-6 Cyber | $20.00 | $2.00 | $100.00 | Cyber specialist (a GPT-6 Astra variant, released 2026-11-05). Approved Daybreak Red customers only; access expanding progressively. Cache writes $25 (30m) / $50 (12h). Not a general-purpose model β timeline | Limited access |
| GPT-6 Astra | $10.00 | $1.00 | $50.00 | Flagship (announced 2026-09-03). Long: $20/$2/$75. Fast mode 2Γ βDeep dive | Legacy |
| GPT-6 Sol | $2.00 | $0.20 | $10.00 | Released 2026-09-22. Successor to GPT-5.6 Sol at 50% off (was $4/$20). 1,050,000 context, 128K output βDeep dive | Legacy |
| GPT-6.1 Sol | $2.00 | $0.10 | $10.00 | Input $2 / output $10 matches GPT-6 Sol, but cached input is $0.10 (5% of input). Cache writes $2.50, Fast 2x, Ultrafast 6x, Batch/Flex 50% off. Prompts above 272,000 input tokens are repriced at 2x input and 1.5x output for the whole request Details β | Legacy |
| GPT-6.1 Terra | $1.00 | $0.05 | $5.00 | Released 2026-10-26. Cache read $0.05 = 5% of input. Re-checked on the OpenAI pricing page (2026-11-10). β details | Current |
| GPT-6 Luna | $0.10 | $0.01 | $0.50 | Released 2026-09-22. Successor to GPT-5.6 Luna at 50% off (was $0.20/$1.20). Free/Go tiers get it in the desktop app βDeep dive | Current |
| GPT-5.6 Sol | $4.00 thru 11/21 | $0.40 | $20.00 | Previous generation of GPT-6 Sol (legacy). Promo (was $5/$30). Long: $8/$0.80/$30 βDeep dive | Legacy |
| GPT-5.6 Terra | $2.00 | $0.20 | $12.00 | Cut 2026-07-30 (was $2.50/$15). Long: $4/$0.40/$18 | Legacy |
| GPT-5.6 Luna | $0.20 | $0.02 | $1.20 | Previous generation of GPT-6 Luna (legacy). Cut 2026-07-30 (was $1/$6). Long: $0.40/$0.04/$1.80 βDeep dive | Legacy |
| GPT-5.5 | $5.00 | $0.50 | $30.00 | Previous flagship | Legacy |
| GPT-5.4 | $2.50 | $0.25 | $15.00 | Balanced (superseded by Terra) | Legacy |
| GPT-5.4 mini | $0.75 | $0.075 | $4.50 | Lightweight tier | Legacy |
| GPT-5.4 nano | $0.20 | $0.02 | $1.25 | Cheapest tier | Legacy |
Primary source: developers.openai.com/api/docs/pricing (2026-08-29 JST). Priority processing renamed to Fast mode. π° GPT-5.6 Sol is at a promo price of $4/$20 (at least through 2026-11-21; was $5/$30). Cloudflare AI Gateway has a limited-time 50% off (through 2026-09-18, Unified Billing only; vs Cloudflare's listed $5/$30 standard). π GPT-5.6 Sol: pricing, promo & competitor comparison β Β· π GPT-5.6 Luna β
Anthropic (Claude)
π Standard API / text input / no cache
| Model | Input | Cache Write 5m | Cache Hit | Output | Notes | Status |
|---|---|---|---|---|---|---|
| Claude Fable 5.2 | $8.00 | $10.00 | $0.20 | $40.00 | Released 2026-11-09 (rates verified 2026-11-12; cache rates verified 2026-11-14). Successor to Fable 5.1 at 20% lower rates ($10/$50 β $8/$40); fast mode $16/$80. Available to all users ("available today on all platforms"). Batch input $4/MTok β previous generation | Current |
| Claude Fable 5.1 | $10.00 | $12.50 | $0.25 | $50.00 | Released 2026-09-01. Fable 5 successor. Cache read 75% off ($1.00β$0.25) βDeep dive | Legacy |
| Claude Fable 5 | $10.00 | $12.50 | $1.00 | $50.00 | Previous-gen flagship. Cache read $1.00 | Legacy |
| Claude Opus 5.5 | $4.00 | $5.00 | $0.20 | $20.00 | Released 2026-09-22. First of the Claude 5.5 family. Cheaper than Opus 5 ($5/$25β$4/$20); cache reads down 60% ($0.50β$0.20). Fast mode $8/$40 βDeep dive | Current |
| Claude Opus 5 | $5.00 | $6.25 | $0.50 | $25.00 | Released 2026-07-24. Same price as Opus 4.8. Fast mode $10/$50. Previous generation to Opus 5.5 | Legacy |
| Claude Opus 4.8 | $5.00 | $6.25 | $0.50 | $25.00 | Previous Opus. Fallback for Claude Max | Legacy |
| Claude Sonnet 5.5 | $2.00 | $2.50 | $0.10 | $10.00 | Released 2026-09-28. Input/output the same as Sonnet 5; cache reads halved from $0.20 to $0.10 (corrected 2026-10-09 against the official page). Details β | Current |
| Claude Sonnet 5 | $2.00 | $2.50 | $0.20 | $10.00 | Released 2026-06-30 (previous generation). $2/$10 now permanent (announced 2026-08-10) | Legacy |
| Claude Sonnet 4.6 | $3.00 | $3.75 | $0.30 | $15.00 | Older balanced model | Legacy |
| Claude Haiku 5.5 | $0.10 | $0.125 | $0.01 | $0.50 | Released 2026-10-07. Fastest and cheapest of the Claude 5.5 family. Above 100,000 tokens: $0.50 input / $2.50 output / $0.05 cache reads / $0.625 (5-min TTL) and $1 (1-hour TTL) cache writes. Batch API is 50% off input and output. The new tokenizer counts the same text as roughly 30% more tokens. Details β | Current |
| Claude Haiku 4.5 | $1.00 | $1.25 | $0.10 | $5.00 | High-volume, classification | Legacy |
Primary source: platform.claude.com (Sonnet 5.5 rows verified against the official Sonnet 5.5 overview and pricing docs via Firecrawl on 2026-10-01; other Anthropic rows retrieved 2026-11-03 JST). Claude Sonnet 5 is now permanently $2/$10 (the $3/$15 increase scheduled for Sep 1 will not occur). Batch 50% off. Fable 5.1 cache read is $0.25 (75% below Fable 5's $1.00). Flat rate full 1M context. π Claude Sonnet 5.5 deep dive β Β· π Claude Fable 5.1 deep dive β
Google (Gemini Developer API)
π Paid tier / Standard / text, image, video input / standard API
| Model | Input | Input>200K | Output | Output>200K | Notes | Status |
|---|---|---|---|---|---|---|
| Gemini 4 Neon (gemini-4-neon) | $3.00 | β | $15.00 | β | Generally available (GA) on 2026-11-18. Flat rate across the full 1M tokens (no tiers). Implicit caching only ($0.15). No free tier. Four thinking levels (low/medium/high/xhigh). β timeline | Current |
| Gemini 4 Hydrogen (gemini-4-hydrogen) | $1.50 | β | $7.50 | β | Announced 2026-11-12. Flat rate across the full 1M tokens (no tiered pricing). Implicit caching only ($0.075). Free tier available, no batch pricing listed. βtimeline | Preview |
| Gemini 4 Carbon (gemini-4-carbon) | $2.00 until 11/30 $4.00(12/1~) | β | $10.00 until 11/30 $20.00(12/1~) | β | Generally available on 2026-11-10 ("available to developers, enterprises, scientists, and consumers"). Implicit caching only ($0.10 until 11/30). Batch API not supported, no free tier. Context length unconfirmed. βtimeline | Current |
| Gemini 4 Argon (gemini-4-argon) | $2.00 until 11/30 $4.00(12/1~) | β | $10.00 until 11/30 $20.00(12/1~) | β | GA in October 2026 (gemini-4-argon). 1M context, 65,536 max output (up to 1,048,576 with long decode). Implicit caching only ($0.10 until 11/30), no free tier. βtimeline | Current |
| Gemini 3.8 Flash | $0.75 until 12/31 $1.50(1/1~) | β | $3.75 until 12/31 $7.50(1/1~) | β | Released 2026-09-02. Best reasoning/coding Flash. 3.8 Flash Cyber is Fairwind-only βDeep dive | Current |
| Gemini 3.7 Flash | $0.75 until 12/31 $1.50(1/1~) | β | $3.75 until 12/31 $7.50(1/1~) | β | Released 2026-08-14. Most capable Flash for coding/agents βDeep dive Retirement date: 2027-01-28 (replacement model gemini-3.8-flash): retirement date and replacement model ID checked in Google Cloud model lifecycle docs (2026-11-10). Not selectable in AI Studio; API requests are auto-routed to 3.8 Flash. | Legacy |
| Gemini 3.6 Flash | $0.75 until 12/31 $1.50(1/1~) | β | $3.75 until 12/31 $7.50(1/1~) | β | Cut to introductory price (was $1.50/$7.50) | Legacy |
| Gemini 3.5 Flash | $1.50 | β | $9.00 | β | GA May 2026 | Legacy |
| Gemini 3.5 Flash-Lite | $0.30 | β | $2.50 | β | GA July 2026. Cheapest Gemini 3.x | Legacy |
| Gemini 3.1 Pro Preview | $2.00 | $4.00 | $12.00 | $18.00 | Preview pricing | Preview |
| Gemini 3.1 Flash-Lite | $0.25 | β | $1.50 | β | Audio $0.50 | Legacy |
| Gemini 2.5 Pro | $1.25 | $2.50 | $10.00 | $15.00 | From Dec 1, 2026, API requests auto-route to Gemini 3.8 Flash and are billed at 3.8 Flash rates | Legacy |
| Gemini 2.5 Flash | $0.30 | β | $2.50 | β | Audio $1.00 From Dec 1, 2026, API requests auto-route to Gemini 3.5 Flash-Lite and are billed at 3.5 Flash-Lite rates | Legacy |
| Gemini 2.5 Flash-Lite | $0.10 | β | $0.40 | β | Audio $0.30 | Legacy |
Gemini 2.5 Pro / 2.5 Flash automatic routing (2026-12-01): the official pricing page states verbatim: "Effective December 1, 2026, all API requests to gemini-2.5-pro will automatically route to Gemini 3.8 Flash (gemini-3.8-flash) and be billed at Gemini 3.8 Flash rates", and "Effective December 1, 2026, all API requests to gemini-2.5-flash will automatically route to Gemini 3.5 Flash-Lite (gemini-3.5-flash-lite) and be billed at Gemini 3.5 Flash-Lite rates". Both are deprecated and no longer selectable in Google AI Studio. Primary source: ai.google.dev/gemini-api/docs/pricing (fetched 2026-11-15 JST).
Gemini 3.7/3.6 Flash introductory price ($0.75/$3.75) valid through Dec 31, 2026; standard price $1.50/$7.50 from Jan 1, 2027. Primary source: ai.google.dev (2026-08-14 JST) Gemini 4 Argon rates also checked on the Gemini 4 Argon model page (2026-11-03 JST).
DeepSeek
π DeepSeek-V4.1-Flash launched 2026-09-10 with across-the-board price cuts. V4 Pro continues after 2026-09-14 β DeepSeek withdrew its 9/14 routing notice
deepseek-flash: lower rates, native vision). V4 Flash and V4 Flash Vision Exp are retired into V4.1 Flash. deepseek-v4-pro, by contrast, keeps running after 2026-09-14 with billing unchanged β the 9/14 routing notice was withdrawn. Peak hours: 01:00-04:00 and 06:00-10:00 UTC, Monday-Friday (JST 10:00-13:00 / 15:00-19:00). Weekends are off-peak all day; off-peak is half the peak rate.| Model | Input (miss) | Input (hit) | Output | Notes | Status |
|---|---|---|---|---|---|
| DeepSeek V4.1 Flash (deepseek-flash) | $0.15 (off-peak) $0.30 (peak) | $0.003 (off-peak) $0.006 (peak) | $0.60 (off-peak) $1.20 (peak) | 1M ctx, 384K out. Native vision (image input). Released 2026-09-10 β guide | Current |
| DeepSeek V4 Pro (-0813) Β· live | $0.66 (off-peak) $1.32 (peak) | $0.022 (off-peak) $0.044 (peak) | $1.98 (off-peak) $3.96 (peak) | These rates are current. DeepSeek states V4 Pro keeps running after 2026-09-14 with billing unchanged (the 9/14 routing notice was withdrawn). No vision; concurrency 500. β continuation details | Current |
β° Peak hours: 01:00-04:00 and 06:00-10:00 UTC, Monday-Friday only (JST 10:00-13:00 / 15:00-19:00; Beijing 09:00-12:00 / 14:00-18:00). Weekends are off-peak all day. Off-peak is half the peak rate. Effective 2026-08-16 16:00 UTC. Old promo rates (V4 Pro $0.435/$0.87, V4 Flash $0.14/$0.28) retired. The legacy model names deepseek-v4-flash and deepseek-v4-flash-vision-exp are retired and served by V4.1 Flash (billed at the Flash price). DeepSeek has not announced a release date for V4.1 Pro. Source: api-docs.deepseek.com (2026-09-13) Β· Change Log (2026-09-10)
xAI (Grok)
π Standard API rate. Fast variant is 2Γ (see official docs for cache/other modes)
| Model | Input | Output | Notes | Status |
|---|---|---|---|---|
| Grok 4.7 | $2.00 | $6.00 | Released 2026-09-21. Fast variant 2Γ ($4/$12) Β· long context $4/$12 above 200K βDeep dive | Current |
| Grok 4.6 | $2.00 | $6.00 | Released 2026-08-12. Fast variant 2Γ ($4/$12) βDeep dive | Legacy |
Primary source: x.ai/news/grok-4-7 Β· docs.x.ai/developers/pricing Β· x.ai/news/grok-4-6 ("Pricing starts at $2 per million input tokens and $6 per million output tokens") Β· x.ai/pricing (consumer/enterprise plans; see docs.x.ai for full API details)
Zhipu (Z.ai / GLM)
π GLM-5.3-Flash at list price (intro 50% promo ended 9/9). GLM-5.3/5.2 at standard pricing
| Model | Input | Cached Input | Output | Notes | Status |
|---|---|---|---|---|---|
| GLM-5.3-Flash | $0.15 | $0.03 | $0.50 | 320B/18B MoE, 1M ctx, 128K out, MIT. Natively multimodal βDeep dive | Current |
| GLM-5.3 | $1.40 | $0.26 | $4.40 | Reasoning model. Cache storage limited-time free | Current |
| GLM-5.2 | $1.40 | $0.26 | $4.40 | MIT license. Cache storage limited-time free | Legacy |
Primary source: docs.z.ai/guides/overview/pricing (2026-09-09) Β· GLM-5.3-Flash announcement. π GLM 5.3 Flash deep dive β
Alibaba (Qwen)
π QwenCloud (international) standard rates. Flagship Qwen3.8-Max and low-cost Qwen3.8-Flash
| Model | Input | Output | Notes | Status |
|---|---|---|---|---|
| Qwen4-Max | $1.50 implicit cache $0.075 | $4.50 | Released 2026-11-15 (current flagship). 1M context, up to 991K input, 131K output (262K when thinking), text/image/video input. Explicit cache writes $1.875 (5 min) / $3.00 (1 hour). QwenCloud pricing | Current |
| Qwen3.8-Flash | $0.15 | $0.47 | Released 2026-08-26. 125B/6B active MoE, 256Kβ1M context, multimodal. Qwen4 preview βDeep dive | Current |
| Qwen3.8-Max | $2.00 | $6.00 | Flagship (2.4T MoE, 1M context). Implicit cache $0.25 | Legacy |
Primary source: qwen.ai/blog (Qwen3.8-Flash-Next) (2026-11-05). Pricing $0.15/$0.47 per the official blog (the official X post lists $0.16/$0.47). π Qwen3.8-Flash deep dive β Β· π¨π³ China AI overview β
Xiaomi (MiMo)
β Checked on the official page (mimo.xiaomi.com/mimo-v2-6) on 2026-11-17. MiMo-V2.6 ships as three variants β Pro, Flash and Pro-UltraSpeed β at the same API prices as V2.5. All three take text, image, video and audio input (text out); cache writes are free for a limited time. Rates below are USD per 1M tokens.
| Model | Input (cache miss) | Input (cache hit) | Output | Notes | Status |
|---|---|---|---|---|---|
| MiMo-V2.6-Pro | $0.435 | $0.0036 | $0.87 | Top of the series. Four input modalities (text, image, video, audio), 1M context. Prices unchanged from V2.5; cache writes free for a limited time βOverview | Current |
| MiMo-V2.6-Flash | $0.14 | $0.0028 | $0.28 | The balanced variant. Same 1M context and four input modalities as Pro. Batch at 50% of the real-time price | Current |
| MiMo-V2.6-Pro-UltraSpeed | $4.35 | $0.036 | $8.70 | Pro weights served faster (up to 20x output speed); 10x the Pro rates | Current |
Sources: mimo.xiaomi.com (MiMo-V2.6) (verified 2026-11-17). Batch is 50% of the real-time price, per the vendor documentation (mimo.mi.com). Weight availability could not be confirmed on the page we retrieved; CNY pricing for China is listed separately. π Chinese model overview β
Moonshot (Kimi)
Verified directly with Firecrawl on platform.kimi.ai (English, USD) and platform.kimi.com (Chinese, CNY) on 2026-11-07. Kimi K3.1 Pro has its own row on the pricing page and has launched (we read it as GA). Figures below are the USD list; the Chinese page shows Β₯13.00/Β₯68.00, a different currency.
| Model | Input | Cached input | Output | Notes | Status |
|---|---|---|---|---|---|
| Kimi K3.1 Pro | $2.00 | $0.20 | $10.00 | 1,048,576 context. Cache billing is on by default (writes $2.00 for 5m, $4.00 for 1h). Batch at 60% of standard. kimi-k3.1-pro βDeep dive | Current |
| Kimi K3.1 Pro Highspeed | $4.00 | $0.20 | $20.00 | Same quality, faster (2x rates). Writes $4.00 (5m) / $8.00 (1h) | Current |
| Kimi K3 | $3.00 | $0.30 | $15.00 | Previous generation. K2 closes to new users from 2026-10-31 and retires end of Dec 2026 | Legacy |
Sources: platform.kimi.ai/docs/pricing/chat (USD) Β· platform.kimi.com/docs/pricing/chat (CNY) (verified 2026-11-07). π Kimi K3.1 Pro deep dive β
Meta (Meta Model API / Muse)
π Standard mode, USD per 1M tokens. Muse Wonder has no Contributor mode and no long-context premium
β Checked the official pricing page (dev.meta.ai / Pricing and rate limits) on 2026-11-05. Muse Wonder (launched 2026-10-12, muse-wonder-1.0) is $4.00 input / $0.20 cached input / $20.00 output. Muse Wonder Full is subscription-only in Muse Code, with no token-metered rate published.
| Model | Input | Cached input | Output | Notes | Status |
|---|---|---|---|---|---|
| Muse Ember (muse-ember-1.0) | $0.10 | Not confirmed | $0.50 | Released 2026-11-09 (rates verified 2026-11-12). Smallest and most efficient Muse model, built for tool calls, routing, extraction and long-document parsing. On the Meta Model API, Muse Code, Google Cloud and Oracle Cloud β timeline | Current |
| Muse Wonder (muse-wonder-1.0) | $4.00 | $0.20 | $20.00 | Launched 2026-10-12. 1M context, no long-context premium. Web search grounding is billed separately at $2.50 per 1,000 queries β timeline | Current |
| Muse Spark (muse-spark-1.3 / 1.2 / 1.1) | $1.25 Contributor $0.10 | $0.15 Contributor $0.002 | $4.25 Contributor $0.20 | Muse Spark 1.1 is Standard mode only. Contributor mode is available via the muse-spark-1.3-contributor / 1.2-contributor variants | Current |
Beyond token pricing: Muse Image (image generation) $0.01 per image including its built-in web and image search, Muse Voice Transcribe $0.18 per hour of audio (billed per second, streaming and non-streaming alike), web search grounding $2.50 per 1,000 queries. Source: dev.meta.ai (Pricing and rate limits), checked 2026-11-05. π Model timeline β
DeepSeek V4.1 Flash β steep price cuts and native vision; V4 Pro continues (9/14 notice withdrawn)
V4.1 Flash costs far less than V4 Pro and adds native image input. It was initially announced that from 2026-09-14 every request β including those to deepseek-v4-pro β would route to V4.1 Flash, but DeepSeek withdrew that notice and confirmed V4 Pro continues. New rates: $0.15/$0.60 off-peak, $0.30/$1.20 peak (cache-miss input / output) β about 77% lower on input, 70% lower on output and 86% lower on cached input versus V4 Pro. Off-peak and weekend runs cut cost further. See the V4.1 Flash guide, V4 Pro continuation and the calculator.
β οΈ Disclaimer
- What βothersβ covers: StepFun (re-verified 2026-11-05: step-5-preview $1.00 / $2.70, step-3.7-flash $0.20 / $1.15) plus smaller vendors whose figures date from 2026-08-04. The major vendors above were re-fetched on 2026-11-05 or 2026-11-03, per the verification dates at the top of this page.
- Pricing may change without notice.
- DeepSeek peak/off-peak pricing takes effect 2026-08-16 16:00 UTC. Weekends are off-peak all day. DeepSeek-V4.1-Flash launched 2026-09-10 with across-the-board price cuts. Requests to V4 Flash / V4 Flash Vision Exp are served by V4.1 Flash at the V4.1 Flash price. V4 Pro, by contrast, continues after 2026-09-14 with billing unchanged (DeepSeek withdrew the 9/14 routing notice).
- This site is informational only.