Model Data Sources
This page tracks the official provider docs used to validate model names and pricing displayed on this site. Benchmark scores are internal editorial scores; provider docs validate model availability, context windows, and token pricing.
Last verified: 2026-09-20| Model | Pricing ($/M input/output) | Notes | Official Sources |
|---|---|---|---|
GPT-6 Astra OpenAI | $10 / $50 | Standard API pricing per MTok below 272K input tokens. Requests above that threshold are $20 input / $75 output per MTok. | |
GPT-Live-1 OpenAI | See pricing notes | Voice sessions cost $0.05 per minute, billed per second. Backend model and tool usage are billed separately. | |
GPT-5.6 Sol OpenAI | $4 / $20 | Promotional API pricing per MTok, available at least through November 21, 2026, per OpenAI. | |
GPT-5.6 Terra OpenAI | $2 / $12 | Standard API price per MTok below 272K input tokens. Above that threshold, pricing is $4 input / $18 output per MTok. | |
GPT-5.6 Luna OpenAI | $0.2 / $1.2 | Standard API price per MTok below 272K input tokens. Above that threshold, pricing is $0.40 input / $1.80 output per MTok. | |
GPT-5.5 OpenAI | $5 / $30 | Standard published rate. | |
Claude Fable 5.1 Anthropic | $10 / $50 | Standard API list price per MTok; prompt-cache reads are $0.25/MTok. | |
Claude Fable 5 Anthropic | $10 / $50 | Standard published rate. | |
Claude Opus 5 Anthropic | $5 / $25 | Standard published rate. | |
Claude Opus 4.8 Anthropic | $5 / $25 | Standard published rate. | |
GPT-5.4 OpenAI | $2.5 / $15 | Standard published rate. | |
Gemini 3.5 Flash Google | $1.5 / $9 | Standard tier; Batch and Flex tiers are lower. | |
Gemini 3.6 Flash Google | $1.5 / $7.5 | Standard paid tier; Batch and Flex tiers are lower. | |
Gemini 3.8 Flash Google | $0.75 / $3.75 | Introductory paid-tier pricing per MTok through December 31, 2026; standard pricing becomes $1.50 input / $7.50 output on January 1, 2027. | |
Gemini 3.8 Live Google | $0.75 / $4.5 | Published text rates per MTok. Audio input is $3/MTok (about $0.005/min) and audio output is $12/MTok (about $0.018/min); image/video input is $1/MTok (about $0.002/min). | |
Gemini 3.8 Live Extended Thinking Google | $0.75 / $4.5 | Published text rates per MTok. Audio input is $3/MTok (about $0.005/min) and audio output is $12/MTok (about $0.018/min); image/video input is $1/MTok (about $0.002/min). | |
Gemini 3.7 Flash Google | $0.75 / $3.75 | Introductory paid-tier pricing per MTok through December 31, 2026; standard pricing becomes $1.50 input / $7.50 output on January 1, 2027. | |
Gemini 3.5 Flash-Lite Google | $0.3 / $2.5 | Standard paid tier; Batch and Flex tiers are lower. | |
Claude Sonnet 5 Anthropic | $2 / $10 | Standard list price per MTok. Anthropic cancelled the previously scheduled September 1, 2026 increase. | |
Claude Haiku 4.5 Anthropic | $1 / $5 | Standard published rate. | |
GPT-5.2-Codex OpenAI | $1.75 / $14 | Standard published rate. | |
Grok 4.5 xAI | $2 / $6 | Standard context pricing. Prompts at or above 200K tokens use xAI’s higher long-context rates. | |
Grok 4.6 xAI | $2 / $6 | Per MTok below 200K prompt tokens; cached input is $0.50/MTok. Prompts at or above 200K use $4 input / $1 cached input / $12 output per MTok. | |
GLM-5.3-Flash Z.ai | $0.15 / $0.5 | Standard Z.ai API list pricing per MTok; cached input is $0.03/MTok. | |
GLM-5.3 Z.ai | $1.4 / $4.4 | Standard Z.ai API pricing per MTok; cached-input pricing is $0.26/MTok. | |
GLM-5 Zhipu AI | $1 / $3.2 | Standard published rate. | |
DeepSeek V4 Pro DeepSeek | $0.66 / $1.98 | Off-peak list price per MTok. DeepSeek continued the dedicated V4 Pro API after September 14; peak rates are $1.32 input / $3.96 output and cache-hit input is $0.022 off-peak or $0.044 peak. Peak hours are 01:00–04:00 and 06:00–10:00 UTC on weekdays. | |
DeepSeek V4.1 Flash DeepSeek | $0.15 / $0.6 | Off-peak list price per MTok. Peak rates are $0.30 input / $1.20 output; cache-hit input is $0.003 off-peak or $0.006 peak. Peak hours are 01:00–04:00 and 06:00–10:00 UTC on weekdays. | |
GPT-5.2 OpenAI | $1.75 / $14 | Standard published rate. | |
Mistral Medium 3.5 Mistral | $1.5 / $7.5 | Standard published rate. | |
Kimi K3 Moonshot AI | $3 / $15 | Kimi lists $0.30/MTok cached input, $3/MTok input, and $15/MTok output. | |
Kimi K2.7 Code Moonshot AI | $0.95 / $4 | Standard Kimi API cache-miss price per MTok; cache-hit input is $0.19/MTok. The higher-throughput variant is $0.38 cached input / $1.90 cache-miss input / $8 output per MTok. | |
Qwen3.8-Flash-Next Alibaba / Qwen | See pricing notes | Open weights for self-hosting; infrastructure costs apply. The separately hosted Qwen3.8-Flash model has its own QwenCloud token pricing. | |
Qwen3.8-Flash Alibaba / Qwen | $0.15 / $0.47 | QwenCloud list prices per MTok; implicit cache reads are $0.016/MTok. | |
Qwen3.8-Omni-Flash Alibaba / Qwen | See pricing notes | The model card directs users to Model Studio pricing; a comparable USD per-token rate was not verified in this pass. | |
Qwen3.8-Max Alibaba / Qwen | See pricing notes | No current public international list price was verified in this pass; check Model Studio before production use. | |
North Small Translate Cohere | See pricing notes | No North-specific production token rate is published. Free-tier Chat V2 access is rate-limited and not for production or commercial use; open weights are CC BY-NC 4.0. Commercial deployment requires a Model Vault license. | |
GPT-OSS-120B OpenAI | $0 / $0 | Free — open weights, self-hosted |