Skip to content

Find the right model.
Know the price.

Every model the gateway serves right now, best first. Prices are in US dollars per 1M tokens, with input, output and cache usage billed separately.

Create an account See plans and allowances

71 models live now

Flagship models 12

The most capable model from each maker, for hard reasoning, coding and agent work.

ModelInputOutputCache readCache writeContext
GPT-6 Astra
OpenAIgpt-6-astra

OpenAI GPT-6 Astra — OpenAI's most capable reasoning model. 1M-token context. Tool calling via /v1/responses; leave temperature at default.

Input$1
Output$5
Cache read$0.1
Cache write$1.25
Context1.1M
Claude Opus 4.8
Anthropicclaude-opus-4-8

Anthropic Claude Opus 4.8 — previous-generation Opus; dependable for coding and analysis. 1M-token context.

Input$0.5
Output$2.50
Cache read$0.05
Cache write$0.625
Context1M
Claude Opus 4.6
Anthropicclaude-opus-4-6

Anthropic Claude Opus 4.6 — previous-generation Opus. 1M-token context.

Input$0.5
Output$2.50
Cache read$0.05
Cache write$0.625
Context1M
GPT-6.1 Sol
OpenAIgpt-6.1-sol

OpenAI GPT-6.1 Sol — reasoning model with tools, vision and structured output. 1M-token context.

Input$0.2
Output$1
Cache read$0.01
Cache write$0.25
Context1.1M
GPT-6 Sol
OpenAIgpt-6-sol

OpenAI GPT-6 Sol — reasoning model with tools, vision and structured output. 1M-token context.

Input$0.2
Output$1
Cache read$0.02
Cache write$0.25
Context1.1M
GPT-5.6 Sol
OpenAIgpt-5.6-sol

OpenAI GPT-5.6 Sol — balanced GPT-5.6 reasoning model; a strong default for Codex and agents. 1M-token context.

Input$0.4
Output$2
Cache read$0.04
Cache write$0.5
Context1.1M
IQuest-Q1Free
IQuestLabiquest-q1

IQuest-Q1 by IQuestLab — free frontier model with top-tier performance that beats DeepSeek V4. Strong at coding, reasoning and agentic tool use, with a 512K-token context window.

InputFree
OutputFree
Cache read—
Cache write—
Context524K
Kimi K3
Moonshotkimi-k3

Moonshot Kimi K3 — open-weight flagship for agentic coding and long context. 1M-token context.

Input$0.3
Output$1.50
Cache read$0.03
Cache write—
Context1M
Qwen3.8 Max
Alibabaqwen3.8-max

Alibaba Qwen3.8 Max flagship

Input$0.2
Output$0.6
Cache read$0.025
Cache write—
Context1M
DeepSeek V4 Pro
DeepSeekdeepseek-v4-pro

DeepSeek V4 Pro — DeepSeek's most capable model for complex reasoning, math and code.

Input$0.035
Output$0.106
Cache read$0.001
Cache write—
Context1M
GLM-5.3
Z.aiglm-5.3

GLM-5.3 — Z.ai latest general-purpose model with reasoning.

Input$0.14
Output$0.44
Cache read$0.026
Cache write—
Context200K
Grok 4.20 Reasoning (Legacy)
xAIgrok-4.20-0309-reasoning
Input$0.125
Output$0.25
Cache read$0.02
Cache write—
Context131K

Fast and low cost 15

Quick replies at a fraction of flagship rates. Good for high volume and simple tasks.

ModelInputOutputCache readCache writeContext
DeepSeek V4.1 Flash
DeepSeekdeepseek-v4.1-flash

DeepSeek V4.1 Flash — latest release (version 1017, October 17, 2026): fast 552B MoE model with native image input and a 1M-token context window.

Input$0.008
Output$0.032
Cache read$0
Cache write—
Context1M
Gemini 3.8 Flash
Googlegemini-3.8-flash

Google Gemini 3.8 Flash — newest fast thinking model with multimodal input. 1M-token context.

Input$0.037
Output$0.188
Cache read$0.004
Cache write$0.002
Context1M
GPT-6 Luna
OpenAIgpt-6-luna

OpenAI GPT-6 Luna — fastest, lowest-cost GPT-6 tier for high-volume work. 1M-token context.

Input$0.01
Output$0.05
Cache read$0.001
Cache write$0.013
Context1.1M
GLM-5.3 Flash
Z.aiglm-5.3-flash

Z.ai GLM-5.3 Flash — fast, very low-cost GLM model.

Input$0.015
Output$0.05
Cache read$0.003
Cache write—
Context1M
MiniMax M3.1 Flash
MiniMaxminimax-m3.1-flash

MiniMax M3.1 Flash — newest fast MiniMax model with tool calling.

Input$0.015
Output$0.06
Cache read$0.002
Cache write—
Context200K
Grok 4.20 Fast
xAIgrok-4.20-fast
Input$0.125
Output$0.25
Cache read$0.02
Cache write—
Context256K
GPT-5.6 Luna
OpenAIgpt-5.6-luna

OpenAI GPT-5.6 Luna — fast, low-cost GPT-5.6 tier. 1M-token context.

Input$0.02
Output$0.12
Cache read$0.002
Cache write$0.025
Context1.1M
gemini 3.6 flash
Googlegemini-3.6-flash
Input$0.075
Output$0.375
Cache read$0.007
Cache write$0.004
Context1M
Gemini Flash Latest
Googlegemini-flash-latest
Input$0.075
Output$0.375
Cache read$0.007
Cache write$0.004
Context200K
gpt-4o-mini
OpenAIgpt-4o-mini

OpenAI GPT-4o mini — small, fast and low-cost.

Input$0.015
Output$0.06
Cache read$0.007
Cache write—
Context128K
GPT-4.1 Mini
OpenAIgpt-4.1-mini

OpenAI GPT-4.1 Mini.

Input$0.04
Output$0.16
Cache read$0.01
Cache write—
Context1M
Grok 4.1 Fast
xAIgrok-4-1-fast

Grok 4.1 fast — balanced speed and quality

Input$0.02
Output$0.05
Cache read$0.005
Cache write—
Context128K
DeepSeek V4 Flash
DeepSeekdeepseek-v4-flash

DeepSeek V4 Flash — fast, low-cost general-purpose model (current generation).

Input$0.008
Output$0.032
Cache read$0
Cache write—
Context200K
Gemini 3.1 Flash Lite
Googlegemini-3.1-flash-lite
Input$0.025
Output$0.15
Cache read$0.003
Cache write$0.008
Context200K
GPT-OSS 120B
OpenAIopenai/gpt-oss-120b

OpenAI gpt-oss-120b — open-weight reasoning model; very low cost.

Input$0.015
Output$0.06
Cache read$0.007
Cache write—
Context131K

More chat models 35

Earlier generations, open-weight and specialist models.

ModelInputOutputCache readCache writeContext
GPT-OSS 20B (NVIDIA)
OpenAIopenai/gpt-oss-20b

OpenAI gpt-oss-20b — small open-weight model; very low cost.

Input$0.002
Output$0.009
Cache read$0.001
Cache write—
Context131K
GPT-6.1 Terra
OpenAIgpt-6.1-terra

OpenAI GPT-6.1 Terra: mid-tier GPT-6.1 model for everyday agentic, coding and writing work, strong with screenshots. Tools, vision and structured output. 1M-token context.

Input$0.1
Output$0.5
Cache read$0.005
Cache write$0.125
Context1.1M
GPT-5.6 Terra
OpenAIgpt-5.6-terra

OpenAI GPT-5.6 Terra — GPT-5.6 reasoning model. 1M-token context.

Input$0.2
Output$1.20
Cache read$0.02
Cache write$0.25
Context1.1M
GPT-5.4
OpenAIgpt-5.4

OpenAI GPT-5.4 — general-purpose model. 1M-token context.

Input$0.25
Output$1.50
Cache read$0.025
Cache write—
Context1.1M
GPT-5.1
OpenAIgpt-5.1

OpenAI GPT-5.1 — previous-generation general-purpose model.

Input$0.125
Output$1
Cache read$0.013
Cache write—
Context272K
gpt-4o
OpenAIgpt-4o

OpenAI GPT-4o — fast multimodal flagship (2024-11-20).

Input$0.25
Output$1
Cache read$0.125
Cache write—
Context128K
GPT-4.1
OpenAIgpt-4.1

OpenAI GPT-4.1 — non-reasoning model with a 1M-token context; fast for simple tasks.

Input$0.2
Output$0.8
Cache read$0.05
Cache write—
Context1M
Claude Opus 4.7
Anthropicclaude-opus-4-7

Anthropic Claude Opus 4.7 — previous-generation Opus. 1M-token context.

Input$0.5
Output$2.50
Cache read$0.05
Cache write$0.625
Context1M
gemini 3 flash
Googlegemini-3-flash
Input$0.05
Output$0.3
Cache read$0.005
Cache write$0.008
Context33K
Gemini 2.5 Flash
Googlegemini-2.5-flash
Input$0.03
Output$0.25
Cache read$0.003
Cache write$0.008
Context200K
Grok 4.20 (Legacy)
xAIgrok-4.20-0309-non-reasoning
Input$0.125
Output$0.25
Cache read$0.02
Cache write—
Context131K
Kimi K2.7 Code
Moonshotkimi-k2.7-code

Moonshot Kimi K2.7 Code — coding-tuned Kimi model. 256K-token context.

Input$0.067
Output$0.335
Cache read$0.015
Cache write—
Context262K
Kimi K2.7
Moonshotkimi-k2.7
Input$0.067
Output$0.335
Cache read$0.015
Cache write—
Context200K
Kimi K2.6
Moonshotkimi-k2.6
Input$0.065
Output$0.341
Cache read$0.015
Cache write—
Context33K
GLM-5.2
Z.aiglm-5.2

GLM-5.2 — Z.ai general-purpose model with reasoning.

Input$0.14
Output$0.44
Cache read$0.026
Cache write—
Context200K
GLM 5.1
Z.aiglm-5.1

GLM-5.1 — Z.ai general-purpose model.

Input$0.14
Output$0.44
Cache read$0.026
Cache write—
Context200K
GLM 5
Z.aiglm-5

GLM-5 — Z.ai flagship general-purpose model.

Input$0.1
Output$0.32
Cache read$0.02
Cache write—
Context33K
GLM 4.7 (dot)
Z.aiglm-4.7

GLM-4.7 — Z.ai general-purpose model.

Input$0.054
Output$0.198
Cache read$0.01
Cache write—
Context200K
MiniMax M3
MiniMaxminimax-m3

MiniMax M3 — low-cost model with a 1M-token context.

Input$0.03
Output$0.12
Cache read$0.006
Cache write—
Context1M
MiniMax M2.7
MiniMaxminimax-m2.7

MiniMax M2.7 — open-weight reasoning model.

Input$0.03
Output$0.12
Cache read$0.006
Cache write—
Context128K
Voxtral Small
Mistralvoxtral-small
Input$0.01
Output$0.03
Cache read$0.001
Cache write—
Context200K
Open Mistral Nemo
Mistralopen-mistral-nemo
Input$0.015
Output$0.015
Cache read$0.002
Cache write—
Context200K
Mistral Tiny
Mistralmistral-tiny
Input$0.015
Output$0.015
Cache read$0.002
Cache write—
Context200K
Ministral 14B
Mistralministral-14b
Input$0.02
Output$0.02
Cache read$0.002
Cache write—
Context200K
Ministral 8B
Mistralministral-8b
Input$0.015
Output$0.015
Cache read$0.002
Cache write—
Context200K
Ministral 3B
Mistralministral-3b
Input$0.01
Output$0.01
Cache read$0.001
Cache write—
Context200K
Codestral
Mistralcodestral
Input$0.03
Output$0.09
Cache read$0.003
Cache write—
Context200K
NVIDIA Riva Translate 4B v2
NVIDIAnvidia/riva-translate-4b-instruct-v2

NVIDIA riva-translate-4b-instruct-v2 — machine translation model.

Input$0.01
Output$0.01
Cache read$0.001
Cache write—
Context131K
NVIDIA Nemotron Safety Guard 8B v3
NVIDIAnvidia/llama-3.1-nemotron-safety-guard-8b-v3

NVIDIA llama-3.1-nemotron-safety-guard-8b-v3 — content-safety / guardrail classifier.

Input$0.005
Output$0.008
Cache read$0.003
Cache write—
Context131K
NVIDIA Nemotron 3.5 Content Safety
NVIDIAnvidia/nemotron-3.5-content-safety

NVIDIA nemotron-3.5-content-safety — content-safety / guardrail classifier.

Input$0.02
Output$0.02
Cache read—
Cache write—
Context131K
NVIDIA Nemotron 3 Ultra 550B
NVIDIAnvidia/nemotron-3-ultra-550b-a55b

NVIDIA nemotron-3-ultra-550b-a55b — open-weight model.

Input$0.06
Output$0.24
Cache read$0.012
Cache write—
Context131K
NVIDIA Nemotron 3 Nano Omni 30B Reasoning
NVIDIAnvidia/nemotron-3-nano-omni-30b-a3b-reasoning

NVIDIA nemotron-3-nano-omni-30b-a3b-reasoning — vision-language model.

Input$0.006
Output$0.024
Cache read—
Cache write—
Context131K
NVIDIA Ising Calibration 1.5 31B
NVIDIAnvidia/ising-calibration-1.5-31b

NVIDIA ising-calibration-1.5-31b — open-weight model.

Input$0.006
Output$0.024
Cache read—
Cache write—
Context131K
Llama 3.2 11B Vision (NVIDIA)
Metameta/llama-3.2-11b-vision-instruct

Meta llama-3.2-11b-vision-instruct — vision-language model.

Input$0.005
Output$0.008
Cache read$0.003
Cache write—
Context131K
Poolside Laguna XS 2.1 (NVIDIA)
Poolsidepoolside/laguna-xs-2.1

Poolside laguna-xs-2.1 — open-weight model.

Input$0.006
Output$0.012
Cache read$0.003
Cache write—
Context131K

Image 2

Image generation and editing.

ModelInputOutputCache readCache writeContext
GPT Image 2.5 Sunburst
OpenAIgpt-image-2.5-sunburst

GPT Image 2.5 Sunburst — precise, high-control image editing (OpenAI).

Input$0.8
Output$3
Cache read$0.2
Cache write—
Context200K
GPT Image 2.5 Flare
OpenAIgpt-image-2.5-flare

GPT Image 2.5 Flare — fast image generation/edit, native multimodal (OpenAI).

Input$0.8
Output$3
Cache read$0.2
Cache write—
Context200K

Embeddings 7

Vectors for search and retrieval. Output is not billed.

ModelInputOutputCache readCache writeContext
Text Embedding 3 Small
OpenAItext-embedding-3-small

OpenAI text-embedding-3-small — 1536-dimension embeddings.

Input$0.002
Output$0
Cache read—
Cache write—
Context8K
Text Embedding 3 Large
OpenAItext-embedding-3-large

OpenAI text-embedding-3-large — 3072-dimension embeddings.

Input$0.013
Output$0
Cache read—
Cache write—
Context200K
Mistral Embed
Mistralmistral-embed
Input$0.01
Output$0
Cache read—
Cache write—
Context200K
Codestral Embed
Mistralcodestral-embed
Input$0.015
Output$0
Cache read—
Cache write—
Context200K
NVIDIA Nemotron-3 Embed 1B
NVIDIAnemotron-3-embed-1b
Input$0.001
Output$0
Cache read—
Cache write—
Context200K
NVIDIA Llama Nemotron VL Embed 1B v2
NVIDIAnemoretriever-vlm-embed
Input$0.001
Output$0
Cache read—
Cache write—
Context200K
bge m3
BAAIbge-m3-q4_k_m
Input$0.001
Output$0
Cache read—
Cache write—
Context8K

Cache rates apply only when the provider reports a cache hit. Plans have their own five-hour allowances and monthly caps. Compare plans. Need a rate for a model marked “Ask us”? Message support.