Models · API pricing
What every major model API costs
82 models with current official list prices from 18 providers, plus 26 older, open-weight or unpriced models — each with a dated reason instead of a blank. Input, output, cached input and batch, per 1M tokens in USD, each with the date we checked and a link to the vendor’s own page. Snapshotted daily from 11 Nov 2026; changes are logged below.
| MODEL | WEIGHTS | INPUT | OUTPUT | CACHED IN | BATCH IN / OUT | CONTEXT | CHECKED | SOURCE | Pack |
|---|---|---|---|---|---|---|---|---|---|
Qwen3.7-MaxAlibaba (Qwen) · qwen3.7-max | Closed | $2.50 | $7.50 | — | — | — | 11 Nov 2026 | Official pageNotesInternational (Singapore) price; cache hits bill at 10% of input. | |
Qwen3.8-MaxAlibaba (Qwen) · qwen3.8-max | Closed | $2.00 | $6.00 | $0.25 | — | 1,000K | 11 Nov 2026 | Official pageNotesSingapore/International scope, flat 0-1M tokens, thinking and non-thinking same price. Implicit cache $0.25; explicit cache creation $2.5, read $0.17. Batch unsupported in Singapore. Global scope (Frankfurt/HK/Tokyo) cheaper: $1.65 in / $4.951 out; US (Virginia) $2/$6. Page shows original pricing excluding limited-time promotions. | |
Qwen4-MaxAlibaba (Qwen) · qwen4-max | Closed | $1.50 | $4.50 | $0.075 | — | — | 19 Nov 2026 | Official pageNotesInternational (Singapore) scope, flat 0-1M tokens, thinking and non-thinking same price. Cache hit (explicit or implicit) $0.075; cache creation $1.875 (5m) / $3 (1h). Global scope is cheaper: $1.413 in / $4.24 out. No Singapore batch price printed. | |
Qwen 3.8 27BAlibaba (Qwen) · qwen3.8-27b | Open | $0.50 | $3.00 | — | — | — | 11 Nov 2026 | Official pageNotesInternational (Singapore) price, 0–1M input tier. | |
Qwen3.7-PlusAlibaba (Qwen) · qwen3.7-plus | Closed | $0.40 | $1.60 | $0.080 | — | 1,000K | 11 Nov 2026 | Official pageNotesSingapore/International. Tiered: 0-256K $0.4 in / $1.6 out; 256K-1M $1.2 in / $4.8 out (thinking same as non-thinking). Pricing page labels these 'List price ... (Limited-time 20% off)' for the alias. Implicit cache $0.08 (<=256K); explicit cache create $0.5, read $0.04. Global scope $0.276 / $1.101 (<=256K). | |
Qwen3.8-FlashAlibaba (Qwen) · qwen3.8-flash | Closed | $0.15 | $0.47 | $0.016 | — | 1,000K | 11 Nov 2026 | Official pageNotesSingapore/International scope, flat 0-1M tokens. Implicit cache $0.016; explicit cache creation $0.2, read $0.016. Batch unsupported in Singapore. Global scope $0.113 in / $0.382 out. | |
Qwen 3.6 Max RetiredAlibaba (Qwen) · qwen3.6-max-previewOnly a preview (qwen3.6-max-preview) was sold; Alibaba retired it on 10 Oct 2026. No price is listed any more. | Closed | Retired 10 Oct 2026 | — | — | — | 11 Nov 2026 | Official page | ||
| Qwen3.5-Max Not on vendor price listAlibaba (Qwen)Not on Alibaba Model Studio's official price list (checked 11 Nov 2026); the page lists qwen3.8-max, qwen3.7-max and qwen-max only. Likely superseded. | Closed | Not on vendor price list | — | — | — | 11 Nov 2026 | Official page | ||
| Qwen3.8-Flash-Next Self-host (open weights)Alibaba (Qwen)Open weights on Hugging Face (qwen-community-1.0). Alibaba's API sells Qwen3.8-Flash, a different model. | Open | No first-party API · weights | — | — | — | 11 Nov 2026 | Official page | ||
| Nova 2 Pro previewAmazon | Closed | $1.25 | $10.00 | — | $0.62 / $5.00 | — | 11 Nov 2026 | Official pageNotesBedrock Standard tier, Global cross-region. In-region/geo us-east-1: $1.375 in / $11.00 out; batch $0.6875 / $5.50. No cache-read price listed. | |
Nova ProAmazon · amazon.nova-pro-v1:0 | Closed | $0.80 | $3.20 | $0.20 | $0.40 / $1.60 | 300K | 11 Nov 2026 | Official pageNotesBedrock Standard tier, us-east-1 in-region. Latency-optimized: $1.00 in / $4.00 out. | |
Nova 2 LiteAmazon · amazon.nova-2-lite-v1:0 | Closed | $0.30 | $2.50 | $0.075 | $0.15 / $1.25 | 1,000K | 11 Nov 2026 | Official pageNotesBedrock Standard tier, Global cross-region inference. Geo cross-region (us./eu./jp. profiles) in us-east-1: $0.33 in / $0.0825 cached / $2.75 out; batch $0.165 / $1.375. Flex and Priority tiers also exist. Web grounding $30 per 1K requests. | |
Nova LiteAmazon · amazon.nova-lite-v1:0 | Closed | $0.060 | $0.24 | $0.015 | $0.030 / $0.12 | 300K | 11 Nov 2026 | Official pageNotesBedrock Standard tier, us-east-1 in-region. | |
Nova MicroAmazon · amazon.nova-micro-v1:0 | Closed | $0.035 | $0.14 | $0.009 | $0.018 / $0.070 | 128K | 11 Nov 2026 | Official pageNotesBedrock Standard tier, us-east-1 in-region. Text-only. | |
| amazon-nova-experimental-chat No public APIAmazonAppears only on the Arena leaderboard; Amazon Bedrock's pricing page lists no such model. | Closed | No public API | — | — | — | 11 Nov 2026 | Official page | ||
Claude Fable 5.1Anthropic · claude-fable-5-1 | Closed | $10.00 | $50.00 | $0.25 | $5.00 / $25.00 | 1,000K | 11 Nov 2026 | Official pageNotesCache writes: 5m $12.50, 1h $20 /MTok. Cache hits 0.025x base input. Full 1M context at standard pricing. | |
Claude Fable 5.2Anthropic · claude-fable-5-2 | Closed | $8.00 | $40.00 | $0.20 | $4.00 / $20.00 | 1,000K | 11 Nov 2026 | Official pageNotesCache writes: 5m $10, 1h $16 /MTok. Cache hits 0.025x base input. Full 1M context at standard pricing. Fast mode $16/$80. Claude Mythos 5.2 (limited availability) shares the same pricing. | |
Claude Opus 5Anthropic · claude-opus-5 | Closed | $5.00 | $25.00 | $0.50 | $2.50 / $12.50 | 1,000K | 11 Nov 2026 | Official pageNotesCache writes: 5m $6.25, 1h $10 /MTok. Full 1M context at standard pricing. Fast mode (research preview) $10/$50. | |
Claude Opus 5.5Anthropic · claude-opus-5-5 | Closed | $4.00 | $20.00 | $0.20 | $2.00 / $10.00 | 1,000K | 11 Nov 2026 | Official pageNotesCache writes: 5m $5, 1h $8 /MTok. Cache hits 0.05x base input. Full 1M context at standard pricing. Fast mode $8/$40. | |
Claude Sonnet 5Anthropic · claude-sonnet-5 | Closed | $2.00 | $10.00 | $0.20 | $1.00 / $5.00 | 1,000K | 11 Nov 2026 | Official pageNotesCache writes: 5m $2.50, 1h $4 /MTok. Introductory $2/$10 pricing is now standard; the planned 2026-09-01 rise to $3/$15 was cancelled. | |
Claude Sonnet 5.5Anthropic · claude-sonnet-5-5 | Closed | $2.00 | $10.00 | $0.10 | $1.00 / $5.00 | 1,000K | 11 Nov 2026 | Official pageNotesCache writes: 5m $2.50, 1h $4 /MTok. Cache hits 0.05x base input. Full 1M context at standard pricing. | |
Claude Haiku 5.5Anthropic · claude-haiku-5-5 | Closed | $0.10 | $0.50 | $0.010 | $0.050 / $0.25 | 1,000K | 11 Nov 2026 | Official pageNotesTiered by prompt length. Up to 100K-token prompts: $0.10 in / $0.50 out, cache hit $0.01, batch $0.05/$0.25. Over 100K: $0.50 in / $2.50 out, cache hit $0.05, batch $0.25/$1.25. Cache writes 5m $0.125 / 1h $0.20 (<=100K). | |
Claude Fable 5 LegacyAnthropic · claude-fable-5Anthropic lists it under “Legacy models (still available)”; price from the current pricing page. | Closed | $10.00 | $50.00 | $1.00 | $5.00 / $25.00 | — | 11 Nov 2026 | Official page | |
Claude Opus 4.5 LegacyAnthropic · claude-opus-4-5-20251101Anthropic lists it under “Legacy models (still available)”; retirement not sooner than 24 Nov 2026. | Closed | $5.00 | $25.00 | $0.50 | $2.50 / $12.50 | — | 11 Nov 2026 | Official page | |
Claude Opus 4.6 LegacyAnthropic · claude-opus-4-6Anthropic lists it under “Legacy models (still available)”. | Closed | $5.00 | $25.00 | $0.50 | $2.50 / $12.50 | — | 11 Nov 2026 | Official page | |
Claude Opus 4.7 LegacyAnthropic · claude-opus-4-7Anthropic lists it under “Legacy models (still available)”. | Closed | $5.00 | $25.00 | $0.50 | $2.50 / $12.50 | — | 11 Nov 2026 | Official page | |
Claude Opus 4.8 LegacyAnthropic · claude-opus-4-8Anthropic lists it under “Legacy models (still available)”. | Closed | $5.00 | $25.00 | $0.50 | $2.50 / $12.50 | — | 11 Nov 2026 | Official page | |
Claude Sonnet 4.6 LegacyAnthropic · claude-sonnet-4-6Anthropic lists it under “Legacy models (still available)”. | Closed | $3.00 | $15.00 | $0.30 | $1.50 / $7.50 | — | 11 Nov 2026 | Official page | |
| ERNIE 5.1 Page blocks automated checksBaiduBaidu Qianfan's pricing pages fail with an SSL error for automated and browser checks (11 Nov). To verify by hand; prices are in CNY. | Closed | Blocked — to verify by hand | — | — | — | 11 Nov 2026 | Official page | ||
Dola Seed 2.0 ProByteDance · seed-2-0-pro-260328 | Closed | $0.50 | $3.00 | $0.10 | — | 256K | 11 Nov 2026 | Official pageNotesStandard, ap-southeast-1. Tiered by prompt length: 0-128K $0.50 in / $3.00 out / $0.10 cached; 128K-256K $1.00 / $6.00 / $0.20. Cache storage $0.0083 per 1M tokens per hour. Flex tier: $0.25 / $1.50 (0-128K), $0.50 / $3.00 (128-256K). Not listed in the Batch table. | |
Dola Seed 2.1 TurboByteDance · dola-seed-2-1-turbo-260628 | Closed | $0.50 | $2.50 | $0.10 | — | 256K | 11 Nov 2026 | Official pageNotesStandard, flat 0-256K. Cache storage $0.0083/1M/hour. Flex: $0.25 in / $0.10 cached / $1.25 out. Not listed in Batch table. | |
Seed 2.0 LiteByteDance · seed-2-0-lite-260428 | Closed | $0.25 | $2.00 | $0.050 | $0.12 / $1.00 | 256K | 11 Nov 2026 | Official pageNotesStandard tiered: 0-128K $0.25/$2.00/$0.05 cached; 128-256K $0.50/$4.00/$0.10. Batch (and Flex) table: $0.125/$1.00 (0-128K). Audio input $3.75 (0-128K). | |
Command RCohere · command-r-08-2024 | Closed | $0.15 | $0.60 | — | — | 128K | 11 Nov 2026 | Official pageNotesLegacy table on same page: Command R 03-2024 $0.50/$1.50 (existing customers). | |
Command R7BCohere · command-r7b-12-2024 | Closed | $0.037 | $0.15 | — | — | 128K | 11 Nov 2026 | Official page | |
Command A LegacyCohere · command-a-03-2025Older model (command-a-03-2025), still live per Cohere's docs but no longer priced on cohere.com/pricing, which now lists Command A+. | Open | No price listed | — | — | 256K | 11 Nov 2026 | Official page | ||
Command A+ Self-host (open weights)Cohere · command-a-plus-05-2026Caveat: Cohere says Command A+ is “free until rate limits are reached”; production use is billed per GPU on Model Vault, not per token, so no per-token price is shown.Cohere: “free until rate limits are reached”; production use runs on its GPU-priced Model Vault. Apache-2.0 weights. | Open | No first-party API · weights · licence | — | — | 128K | 11 Nov 2026 | Official pageNotesCohere's pricing card shows 'Free' (free API key and free weights download). Per-token production API pricing is not published, so we show no price. | ||
DeepSeek V4 ProDeepSeek · deepseek-v4-pro | Open | $1.32 | $3.96 | $0.044 | — | 1,000K | 11 Nov 2026 | Official pageNotesPeak-rate list price shown in fields. Off-peak is half: $0.66 in / $1.98 out / $0.022 cache hit. No batch pricing listed. | |
DeepSeek V4.1 FlashDeepSeek · deepseek-flash | Open | $0.30 | $1.20 | $0.006 | — | 1,000K | 11 Nov 2026 | Official pageNotesPeak-rate list price shown in fields. Off-peak (outside 01:00-04:00 and 06:00-10:00 UTC Mon-Fri) is half: $0.15 in / $0.6 out / $0.003 cache hit. No batch pricing listed. | |
DeepSeek V4 Flash RetiredDeepSeek · deepseek-v4-flashDeepSeek replaced it on 10 Sep 2026; the old ID is routed to V4.1 Flash and billed at that price. | Open | Retired 10 Sep 2026 | — | — | — | 11 Nov 2026 | Official page | ||
Gemini 4 NeonGoogle · gemini-4-neon | Closed | $3.00 | $15.00 | $0.15 | $1.50 / $7.50 | — | 20 Nov 2026 | Official pageNotesStandard paid tier. Batch $1.50 in / $7.50 out / $0.075 cache. Flex $1.50 / $7.50. Priority $5.40 / $27.00. Implicit caching only. Not on Free tier. No scheduled price change is printed for this model. | |
Gemini 3.1 Pro previewGoogle · gemini-3.1-pro-preview | Closed | $2.00 | $12.00 | $0.20 | $1.00 / $6.00 | 1,049K | 11 Nov 2026 | Official pageNotesTiered: prompts <=200K $2.00 in / $12.00 out; >200K $4.00 / $18.00. Cache $0.20/$0.40 + storage $4.50 per 1M tokens/hour. Batch $1.00/$6.00 (<=200K), $2.00/$9.00 (>200K). No Free tier. | |
Gemini 4 Argon CHANGE ANNOUNCEDGoogle · gemini-4-argon | Closed | $2.00 | $10.00 | $0.10 | $1.00 / $5.00 | 1,049K | 11 Nov 2026 | Official pageNotesStandard paid tier. Price change announced for 2026-12-01. Some of the new prices are not in the Google quote we saved, so they are not shown. Priority $3.60/$18.00 (the later price is not in the Google quote we saved). Invite-only Contributor tier $1.00/$5.00 (data used to improve products). Not on Free tier. | |
Gemini 4 Carbon CHANGE ANNOUNCEDGoogle · gemini-4-carbon | Closed | $2.00 | $10.00 | $0.10 | $1.00 / $5.00 | 1,049K | 11 Nov 2026 | Official pageNotesStandard paid tier. Price change announced for 2026-12-01. Some of the new prices are not in the Google quote we saved, so they are not shown. Priority tier $3.60/$18.00 (the later price is not in the Google quote we saved). Thinking tokens billed as output. Implicit caching only. Not on Free tier. | |
Gemini 4 HydrogenGoogle · gemini-4-hydrogen | Closed | $1.50 | $7.50 | $0.075 | — | — | 13 Nov 2026 | Official pageNotesVertex AI global endpoint, text, base tier. Non-global endpoints are 10% higher ($1.65 in / $8.25 out). Google AI Studio's own page lists the model but its price table could not be read automatically. | |
Gemini 3.6 Flash CHANGE ANNOUNCEDGoogle · gemini-3.6-flash | Closed | $0.75 | $3.75 | $0.075 | $0.38 / $1.88 | 1,049K | 11 Nov 2026 | Official pageNotesPrice change announced for 2027-01-01. Some of the new prices are not in the Google quote we saved, so they are not shown. Cache storage $0.50 (the later price is not in the Google quote we saved) per 1M tokens/hour. Free tier available. | |
Gemini 3.8 Flash CHANGE ANNOUNCEDGoogle · gemini-3.8-flash | Closed | $0.75 | $3.75 | $0.075 | $0.38 / $1.88 | 1,049K | 11 Nov 2026 | Official pageNotesPrice change announced for 2027-01-01. Some of the new prices are not in the Google quote we saved, so they are not shown. Cache storage $0.50 (the later price is not in the Google quote we saved) per 1M tokens/hour. Priority $1.35/$6.75. Free tier available. | |
Gemini 3 FlashGoogle · gemini-3-flash-preview | Closed | $0.50 | $3.00 | $0.050 | $0.25 / $1.50 | — | 11 Nov 2026 | Official pageNotesPreview model (gemini-3-flash-preview). | |
Gemini 3.5 Flash-LiteGoogle · gemini-3.5-flash-lite | Closed | $0.30 | $2.50 | $0.030 | $0.15 / $1.25 | 1,049K | 11 Nov 2026 | Official pageNotesFlat price (text/image/video/audio input $0.30). Cache storage $1.00 per 1M tokens/hour. Flex cache $0.02. Priority $0.54/$4.50. | |
Gemini 3.1 Flash-LiteGoogle · gemini-3.1-flash-lite | Closed | $0.25 | $1.50 | $0.025 | $0.12 / $0.75 | 1,049K | 11 Nov 2026 | Official pageNotesAudio input $0.50; audio cache $0.05. Cache storage $1.00 per 1M tokens/hour. Priority $0.45/$2.70. | |
Gemini 3.5 Flash LegacyGoogle · gemini-3.5-flashDeprecated; the endpoint auto-routes to Gemini 3.6 Flash. Price from Google's Vertex AI page (global). | Closed | $1.50 | $9.00 | $0.15 | — | — | 11 Nov 2026 | Official page | |
Gemini 2.5 Pro LegacyGoogle · gemini-2.5-proDeprecated: shuts down 1 Dec 2026 and then routes to Gemini 3.8 Flash. Price for prompts up to 200K tokens. | Closed | $1.25 | $10.00 | $0.12 | $0.62 / $5.00 | — | 11 Nov 2026 | Official page | |
Gemini 3.7 Flash LegacyGoogle · gemini-3.7-flashDeprecated; the endpoint auto-routes to Gemini 3.8 Flash. Vertex AI introductory price until 31 Dec 2026, then $1.50 / $7.50. | Closed | $0.75 | $3.75 | $0.075 | — | — | 11 Nov 2026 | Official page | |
Gemini 3 Pro RetiredGoogle · gemini-3-pro-previewGoogle shut down gemini-3-pro-preview on 9 Mar 2026; no price is listed any more. | Closed | Retired 9 Mar 2026 | — | — | — | 11 Nov 2026 | Official page | ||
Muse Wonder previewMeta · muse-wonder-1.0 | Closed | $4.00 | $20.00 | $0.20 | — | 1,049K | 11 Nov 2026 | Official pageNotesStandard mode only (no Contributor discount). Web search grounding extra: $2.50 per 1,000 queries. No batch pricing listed. | |
| Muse SparkMeta | Closed | $1.25 | $4.25 | — | — | — | 11 Nov 2026 | Official pageNotesArena entry is unversioned; every listed Muse Spark version has this price | |
Muse Spark 1.1Meta · muse-spark-1.1 | Closed | $1.25 | $4.25 | $0.15 | — | — | 11 Nov 2026 | Official pageNotesStandard mode price. | |
Muse Spark 1.2Meta · muse-spark-1.2 | Closed | $1.25 | $4.25 | $0.15 | — | — | 11 Nov 2026 | Official pageNotesStandard mode price. | |
Muse Spark 1.3Meta · muse-spark-1.3 | Closed | $1.25 | $4.25 | $0.15 | — | 1,049K | 11 Nov 2026 | Official pageNotesContributor mode (muse-spark-1.3-contributor; content may be used to train Meta models): $0.10 in / $0.002 cached / $0.20 out. Same standard price for muse-spark-1.2. Grounding $2.50 per 1,000 queries. | |
Muse EmberMeta · muse-ember-1.0 | Closed | $0.10 | $0.50 | $0.004 | — | 1,049K | 11 Nov 2026 | Official pageNotesRequests over 200K input tokens (cached counted): $0.30 in / $0.012 cached / $1.50 out, applied to all tokens. Standard mode only. Grounding $2.50 per 1,000 queries. | |
| Muse Spark Open Self-host (open weights)MetaMeta lists it under open-weight models only (gated download, Muse Model License); not in the Model API price table. | Open | No first-party API · weights · licence | — | — | 1,049K | 11 Nov 2026 | Official page | ||
| Muse Wonder Full No public APIMetaMeta: Wonder Full is available only in Muse Code; the API sells muse-wonder-1.0. | Closed | No public API | — | — | — | 11 Nov 2026 | Official page | ||
MiniMax-M3MiniMax · MiniMax-M3 | Open | $0.30 | $1.20 | $0.060 | — | 1,000K | 11 Nov 2026 | Official pageNotesPrices shown are after 'Permanent 50% off' (struck-through list $0.60 / $2.40 / $0.12). Tiered: <=512K input $0.30/$1.20; >512K input $0.60/$2.40 (cache $0.12). Priority tier (service_tier=priority) is 1.5x standard: $0.45/$1.80 (<=512K). | |
MiniMax-M3.1-FlashMiniMax · MiniMax-M3.1-FlashCaveat: Arena lists “MiniMax-M3.1”; MiniMax’s official pages list only M3.1-Flash, so we match them. Check the model ID before you rely on it. | Open | $0.15 | $0.60 | $0.015 | — | 1,000K | 11 Nov 2026 | Official pageNotesFlat price, no context tiers listed. No batch pricing listed. | |
Mistral Large 4Mistral · mistral-large-4Caveat: Mistral’s pricing page still says $0.50 / $1.50 for “Mistral Large”; the Large 4 model page says $0.70 / $2.10. We show the model page’s price. | Open | $0.70 | $2.10 | $0.070 | — | 1,000K | 11 Nov 2026 | Official pageNotesmistral.ai/pricing states batch processing 'reduces the price by 50%'; no per-model batch figure shown. | |
Codestral (25.08)Mistral · codestral-2508 | Closed | $0.30 | $0.90 | — | — | 128K | 11 Nov 2026 | Official page | |
Mistral Small 4Mistral · mistral-small-2603 | Open | $0.15 | $0.60 | — | — | 256K | 11 Nov 2026 | Official pageNotesCached/batch not shown on model card; general pricing page cites 50% batch discount. | |
Mistral Medium 3.5 LegacyMistral · mistral-medium-3-5Deprecated 31 Oct 2026 (replacement: Mistral Large 4). Mistral no longer shows its price on either pricing page. | Open | No price listed | — | — | 256K | 11 Nov 2026 | Official page | ||
Kimi K3Moonshot AI · kimi-k3 | Open | $3.00 | $15.00 | $0.30 | $1.80 / $9.00 | 1,049K | 11 Nov 2026 | Official pageNotesCache writes: 5m TTL $3.00, 1h TTL $6.00 per MTok. Batch = 60% of standard (batch cached $0.18). Prices exclude taxes. | |
Kimi K3.1 ProMoonshot AI · kimi-k3.1-pro | Open | $2.00 | $10.00 | $0.20 | $1.20 / $6.00 | 1,049K | 11 Nov 2026 | Official pageNotesCache writes billed separately: 5m TTL $2.00, 1h TTL $4.00 per MTok. kimi-k3.1-pro-highspeed: $4.00 in / $20.00 out / $0.20 cached. Batch = 60% of standard (batch cached $0.12). Prices exclude taxes. | |
Kimi K2.7 Code deprecatedMoonshot AI · kimi-k2.7-code | Closed | $0.95 | $4.00 | $0.19 | $0.57 / $2.40 | 262K | 11 Nov 2026 | Official pageNoteskimi-k2.7-code-highspeed: $1.90 in / $8.00 out / $0.38 cached. Batch cached $0.114. | |
Kimi K2.6 LegacyMoonshot AI · kimi-k2.6K2 series closed to new users from 31 Oct 2026; Moonshot plans to retire all K2 models by end of Dec 2026. | Open | $0.95 | $4.00 | $0.16 | — | — | 11 Nov 2026 | Official page | |
GPT-5.4 ProOpenAI · gpt-5.4-pro-2026-03-05 | Closed | $30.00 | $180.00 | — | $15.00 / $90.00 | — | 11 Nov 2026 | Official pageNotesPrompts over 272K tokens cost $60 in / $270 out. | |
GPT-5.5 ProOpenAI · gpt-5.5-pro | Closed | $30.00 | $180.00 | — | $15.00 / $90.00 | 1,050K | 11 Nov 2026 | Official pageNotesShort context <=272K input tokens; long context >272K: $60 in / $270 out. No cached-input rate. | |
GPT-6 AstraOpenAI · gpt-6-astra | Closed | $10.00 | $50.00 | $1.00 | $5.00 / $25.00 | 1,050K | 11 Nov 2026 | Official pageNotesShort context <=272K input tokens; long context >272K: $20 in / $2 cached / $75 out; batch long $10/$37.50. Cache writes $12.50 (30m) / $25 (12h). Fast mode 2x. | |
GPT-6.2 AstraOpenAI · gpt-6.2-astra | Closed | $8.00 | $40.00 | $0.40 | $4.00 / $20.00 | — | 19 Nov 2026 | Official pageNotesStandard tier, short context (<=272K input tokens). Batch short context $4 in / $20 out / $0.20 cached. Cache writes $10 (30m) / $20 (12h). The long-context (>272K) table is not in the cached copy of the page, so it is not recorded here. | |
GPT-5.5OpenAI · gpt-5.5 | Closed | $5.00 | $30.00 | $0.50 | $2.50 / $15.00 | 1,050K | 11 Nov 2026 | Official pageNotesShort context <=272K input tokens; long context >272K: $10 in / $1 cached / $45 out; batch long $5/$22.50. | |
GPT-5.6 Sol CHANGE ANNOUNCEDOpenAI · gpt-5.6-solCaveat: OpenAI marks this price as promotional until 30 Nov 2026; the price after that is not stated. | Closed | $4.00 | $20.00 | $0.40 | $2.00 / $10.00 | 1,050K | 11 Nov 2026 | Official pageNotesShort context <=272K input tokens; long context >272K: $8 in / $0.80 cached / $30 out; batch long $4/$15. PROMOTIONAL pricing until 2026-11-30; post-promo price not stated. | |
GPT-5.4OpenAI · gpt-5.4 | Closed | $2.50 | $15.00 | $0.25 | $1.25 / $7.50 | 1,050K | 11 Nov 2026 | Official pageNotesShort context <=272K input tokens; long context >272K: $5 in / $0.50 cached / $22.50 out; batch long $2.50/$11.25. | |
GPT-5.6 TerraOpenAI · gpt-5.6-terra | Closed | $2.00 | $12.00 | $0.20 | $1.00 / $6.00 | 1,050K | 11 Nov 2026 | Official pageNotesShort context <=272K input tokens; long context >272K: $4 in / $0.40 cached / $18 out; batch long $2/$9. | |
GPT-6 SolOpenAI · gpt-6-sol | Closed | $2.00 | $10.00 | $0.20 | $1.00 / $5.00 | 1,050K | 11 Nov 2026 | Official pageNotesShort context <=272K input tokens; long context >272K: $4 in / $0.40 cached / $15 out; batch long $2/$7.50. | |
GPT-6.1 SolOpenAI · gpt-6.1-sol | Closed | $2.00 | $10.00 | $0.10 | $1.00 / $5.00 | 1,050K | 11 Nov 2026 | Official pageNotesShort context <=272K input tokens; long context >272K: $4 in / $0.20 cached / $15 out; batch long $2/$7.50. Cache writes $2.50 (30m) / $5 (12h). Fast 2x, Ultrafast 6x. | |
GPT-6.2 SolOpenAI · gpt-6.2-sol | Closed | $2.00 | $10.00 | $0.10 | $1.00 / $5.00 | — | 20 Nov 2026 | Official pageNotesStandard tier, short context (<=272K input tokens). Batch short context $1 in / $5 out / $0.05 cached. Cache writes $2.50 (30m) / $5 (12h). The long-context (>272K) table is not read here, so it is not recorded. | |
GPT-6.1 TerraOpenAI · gpt-6.1-terra | Closed | $1.00 | $5.00 | $0.050 | $0.50 / $2.50 | 1,050K | 11 Nov 2026 | Official pageNotesShort context <=272K input tokens; long context >272K: $2 in / $0.10 cached / $7.50 out; batch long $1/$3.75. Cache writes $1.25 (30m) / $2.50 (12h). | |
GPT-5.4 miniOpenAI · gpt-5.4-mini | Closed | $0.75 | $4.50 | $0.075 | $0.38 / $2.25 | — | 11 Nov 2026 | Official pageNotesNo long-context tier listed. | |
GPT-5.6 LunaOpenAI · gpt-5.6-luna | Closed | $0.20 | $1.20 | $0.020 | $0.10 / $0.60 | 1,050K | 11 Nov 2026 | Official pageNotesShort context <=272K input tokens; long context >272K: $0.40 in / $0.04 cached / $1.80 out; batch long $0.20/$0.90. | |
GPT-6 LunaOpenAI · gpt-6-luna | Closed | $0.10 | $0.50 | $0.010 | $0.050 / $0.25 | 1,050K | 11 Nov 2026 | Official pageNotesShort context <=272K input tokens; long context >272K: $0.20 in / $0.02 cached / $0.75 out; batch long $0.10/$0.375. | |
GPT-5.2 Pro LegacyOpenAI · gpt-5.2-proOlder model still listed on OpenAI's pricing page; no cached-input price. | Closed | $21.00 | $168.00 | — | $10.50 / $84.00 | — | 11 Nov 2026 | Official page | |
GPT-5.2 LegacyOpenAI · gpt-5.2Older model still listed on OpenAI's pricing page; superseded by GPT-5.4 and later. | Closed | $1.75 | $14.00 | $0.17 | $0.88 / $7.00 | — | 11 Nov 2026 | Official page | |
Step 5StepFun · step-5-preview | Closed | $1.00 | $2.70 | $0.10 | — | — | 11 Nov 2026 | Official pageNotesStepFun's international page lists it as step-5-preview. | |
Hy4 CHANGE ANNOUNCEDTencent · hy4Caveat: Tencent’s price table does not label its region (likely Singapore/international). A 50% promotional rate applies until 31 Dec 2026. | Closed | $0.83 | $2.50 | $0.042 | — | — | 11 Nov 2026 | Official pageNotesTokenHub list price, USD per 1M (Singapore/intl table). Promotion: until 2026-12-31 23:59:59 Beijing time, Hy4 inference settled at 50% of list price. Token Plan subscriptions sold separately. | |
Hy4 Preview previewTencent · hy4-previewCaveat: Tencent’s price table does not label its region (likely Singapore/international). A 50% promotional rate applies until 31 Dec 2026. | Open | $0.83 | $2.50 | $0.042 | — | 1,000K | 11 Nov 2026 | Official pageNotesSame list price as hy4; same 50%-off promotion through 2026-12-31 (Beijing time). | |
Hy3Tencent · hy3Caveat: Tencent’s price table does not label its region (likely Singapore/international). | Closed | $0.13 | $0.53 | $0.033 | — | — | 11 Nov 2026 | Official pageNotesTokenHub list price, USD per 1M. | |
MiMo-V2.6-ProXiaomi · mimo-v2.6-pro | Open | $0.43 | $0.87 | $0.004 | — | 1,000K | 11 Nov 2026 | Official pageNotesInternational USD card on mimo.mi.com. China (CNY): ¥3 in / ¥6 out / ¥0.025 cache hit per 1M. MiMo-V2.6-Pro-UltraSpeed (up to 20x output speed): $4.35 in / $8.7 out / $0.036 cache hit. Batch price not shown on pages read. Token Plan subscription sold separately. | |
MiMo-V2.6-FlashXiaomi · mimo-v2.6-flash | Open | $0.14 | $0.28 | $0.003 | — | 1,000K | 11 Nov 2026 | Official pageNotesInternational USD card on mimo.mi.com. China (CNY): ¥1 in / ¥2 out / ¥0.02 cache hit per 1M. Batch price not shown on pages read. | |
MiMo-V2.5-Pro RetiredXiaomi · mimo-v2.5-proXiaomi retired it on 21 Oct 2026; replaced by MiMo-V2.6-Pro. No price is listed any more. | Open | Retired 21 Oct 2026 | — | — | — | 11 Nov 2026 | Official page | ||
GLM-5.1Z.ai · glm-5.1 | Open | $1.40 | $4.40 | $0.26 | — | — | 11 Nov 2026 | Official pageNotesListed as a current text model on Z.ai. | |
GLM-5.2Z.ai · glm-5.2 | Open | $1.40 | $4.40 | $0.26 | — | — | 11 Nov 2026 | Official pageNotesListed as a current text model on Z.ai. | |
GLM-5.3Z.ai · glm-5.3 | Open | $1.40 | $4.40 | $0.26 | — | 1,000K | 11 Nov 2026 | Official pageNotesCached input storage 'Limited-time Free'. No batch pricing on page. Same price as GLM-5.2 and GLM-5.1. | |
GLM-5.3-FlashZ.ai · glm-5.3-flash | Open | $0.15 | $0.50 | $0.030 | — | 1,000K | 11 Nov 2026 | Official pageNotesCached input storage 'Limited-time Free'. GLM-5.3-FlashX (faster, ~200 tok/s): $0.37 in / $0.075 cached / $1.25 out. No batch pricing. | |
| Inkling 2 Smallthinky | Closed | $0.30 | $1.00 | $0.030 | — | — | 11 Nov 2026 | Official pageNotesTinker serverless inference (beta), prices as of 19 Oct 2026, 256K context. | |
| U2 Flash Region-onlyunisoundPriced in RMB only on Unisound's MaaS page; a limited-time 60%-off badge is shown, so these may be discounted prices. | Closed | ¥0.6 / ¥1.2 CNY (≈$0.090 / $0.180) | — | — | — | 11 Nov 2026 | Official page | ||
Grok 4.5xAI · grok-4.5 | Closed | $2.00 | $6.00 | $0.30 | — | 500K | 11 Nov 2026 | Official pageNotesPrompts >=200K tokens: $4.00 in / $0.60 cached / $12.00 out. | |
Grok 4.6xAI · grok-4.6 | Closed | $2.00 | $6.00 | $0.50 | — | 500K | 11 Nov 2026 | Official pageNotesPrompts >=200K tokens: $4.00 in / $1.00 cached / $12.00 out. | |
Grok 4.7xAI · grok-4.7 | Closed | $2.00 | $6.00 | $0.50 | — | 500K | 11 Nov 2026 | Official pageNotesPrompts >=200K tokens: $4.00 in / $1.00 cached / $12.00 out (higher rate applies to all tokens once prompt reaches 200K). Model page says Batch API billed at a 20% discount to standard rates; no separate batch figures shown. | |
Grok 4.20xAI · grok-4.20-0309-reasoning | Closed | $1.25 | $2.50 | $0.20 | — | 1,000K | 11 Nov 2026 | Official pageNotesPrompts >=200K tokens: $2.50 in / $0.40 cached / $5.00 out. Same price for grok-4.20-0309-non-reasoning and grok-4.20-multi-agent-0309. | |
Grok 4.20 Multi-AgentxAI · grok-4.20-multi-agent-0309 | Closed | $1.25 | $2.50 | $0.20 | — | — | 11 Nov 2026 | Official pageNotesPrompts of 200K+ tokens bill $2.50 in / $5.00 out for the whole request. | |
Grok 4.3xAI · grok-4.3 | Closed | $1.25 | $2.50 | $0.20 | — | 1,000K | 11 Nov 2026 | Official pageNotesPrompts >=200K tokens: $2.50 in / $0.40 cached / $5.00 out. | |
Grok Build 0.1 previewxAI · grok-build-0.1 | Closed | $1.00 | $2.00 | $0.20 | — | 256K | 11 Nov 2026 | Official pageNotesPrompts >=200K tokens: $2.00 in / $0.40 cached / $4.00 out. | |
Source: official vendor pricing pages (linked per row). USD per 1M tokens, base tier, standard (not priority/flex) service. Every row has a price or a reason chip with the date we checked and the source: Legacy 14 · Retired 4 · No public API 2 · Self-host (open weights) 3 · Region-only 1 · Page blocks automated checks 1 · Not on vendor price list 1. “Input under $” only counts current list prices. “—” in a single column means the vendor does not publish that particular price (e.g. no batch tier). JSON · CSV
Announced price changes
Changes vendors have published for a future date. We record them as the vendor states them.
| EFFECTIVE | MODEL | INPUT | OUTPUT | NOTE | Pack |
|---|---|---|---|---|---|
| 30 Nov 2026 | GPT-5.6 SolOpenAI | $4.00 → — | $20.00 → — | OpenAI marks the current price promotional until 30 Nov; the price after that is not stated. Source | |
| 1 Dec 2026 | Gemini 4 ArgonGoogle | $2.00 → $4.00 | $10.00 → not in the Google quote we saved | Google's pricing page: standard paid tier doubles. Source | |
| 1 Dec 2026 | Gemini 4 CarbonGoogle | $2.00 → $4.00 | $10.00 → not in the Google quote we saved | Google's pricing page: standard paid tier doubles. Source | |
| 31 Dec 2026 | Hy4Tencent | $0.83 → not in the Tencent quote we saved | $2.50 → not in the Tencent quote we saved | List price shown; a 50% promotional settlement runs until 31 Dec 2026 (Beijing time). Region is not labelled on the page (likely Singapore/international). Source | |
| 1 Jan 2027 | Gemini 3.6 FlashGoogle | $0.75 → $1.50 | $3.75 → not in the Google quote we saved | Google's pricing page: same schedule as 3.8 Flash. Source | |
| 1 Jan 2027 | Gemini 3.8 FlashGoogle | $0.75 → $1.50 | $3.75 → not in the Google quote we saved | Google's pricing page: price doubles. Source |
Source: official vendor pricing pages, checked 22 Nov 2026. Share card: 1200×630 · 1080×1350
Price history
Source: official vendor pricing pages (linked per row) · daily snapshots since 11 Nov 2026 · CSV · One point per day we checked; history starts 11 Nov 2026.
Price-change log
Every change between two daily snapshots: date, old → new, source.
| DATE | MODEL | FIELD | CHANGE | SOURCE | Pack |
|---|---|---|---|---|---|
| No changes recorded since tracking began on 11 Nov 2026. | |||||
Cost for your usage
Base-tier list prices. Rupees at USD/INR 92.6752 (ECB reference rates, 20 Nov 2026); GST not included.