Models

Input

$0.438

Output

$1.314

Unit

per 1M tokens

Cache read

$0.0219

Context

1M

Max output

64K

Capability

—

First token

—

30-day uptime

—

Tencent's Hy4 flagship reasoning model with native vision for agentic coding, computer use, finance, and research workflows.

Input

Free

Output

Free

Unit

per 1M tokens

Cache read

Free

Context

128K

Max output

4K

Capability

—

First token

—

30-day uptime

—

Free deterministic test model answered by the gateway itself. Replies with fixed text, or calls the first declared tool, on every protocol, streaming or not. Use it to verify an integration before switching to a real model.

Input

from $0.90list $1.00

Output

from $4.50list $5.00

Discount

-10%

Unit

per 1M tokens

Cache read

from $0.045list $0.05

Context

1.1M

Max output

128K

Capability

168

First token

—

30-day uptime

—

Providers

lowest of 2 providers

OpenAI's fast, efficient GPT-6.1 model for visual reasoning, precise computer use, and clear, natural writing.

Input

$2.0878

Output

$10.9208

Unit

per 1M tokens

Cache read

$0.20878

Context

1M

Max output

1M

Capability

—

First token

—

30-day uptime

—

Providers

lowest of 2 providers

Moonshot AI's flagship 3T-class open-weight model with 1M context, strong coding, 3D creation, and financial research.

Input

$4.1756

Output

$21.8416

Unit

per 1M tokens

Cache read

$0.20878

Context

1M

Max output

1M

Capability

—

First token

—

30-day uptime

—

Providers

lowest of 2 providers

Kimi K3.1 Pro served at 2-3x output speed for latency-sensitive workloads.

Input

from $0.015list $0.30

Output

from $0.06list $1.20

Discount

-95%

Unit

per 1M tokens

Cache read

from $0.0003list $0.006

Context

1M

Max output

384K

Capability

159

First token

—

30-day uptime

—

Providers

lowest of 8 providers

DeepSeek V4.1 Flash 1017: natively multimodal vision understanding, with stronger coding and agentic performance than the 0910 release, 1M context and 384K max output.

Input

$0.165

Output

$0.55

Unit

per 1M tokens

Cache read

$0.022

Context

1M

Max output

262K

Capability

—

First token

—

30-day uptime

—

NVIDIA's open 120B vision-language MoE for agentic tool use, coding, and document understanding.

Input

$1.80list $2.00

Output

$9.00list $10.00

Discount

-10%

Unit

per 1M tokens

Cache read

$0.09list $0.10

Context

1M

Max output

66K

Capability

166

First token

—

30-day uptime

—

Providers

lowest of 2 providers

Google's flagship reasoning model for complex software engineering, deep mathematical and scientific problem solving, autonomous multi-step agents, and long decode up to 1M tokens.

Input

from $0.11

Output

from $0.55

Unit

per 1M tokens

Cache read

from $0.011

Context

1M

Max output

128K

Capability

160

First token

—

30-day uptime

—

Providers

lowest of 3 providers

Anthropic's fastest and most affordable small model for high-volume, latency-sensitive work such as classification, routing, extraction, and subagents.

Input

$1.35list $1.50

Output

$6.75list $7.50

Discount

-10%

Unit

per 1M tokens

Cache read

from $0.135list $0.15

Context

131K

Max output

33K

Capability

—

First token

—

30-day uptime

—

Providers

lowest of 2 providers

An update to Nano Banana 2 for image generation and conversational editing, with improved visual quality, multi-turn character consistency, accurate text rendering, and search-grounded generation at 1K, 2K, and 4K.

Input

$0.77

Output

$2.31

Unit

per 1M tokens

Cache read

$0.077

Context

1M

Max output

262K

Capability

151

First token

—

30-day uptime

—

Mistral AI's 1.05T-parameter multimodal MoE for reasoning, coding, and agentic work.

Embedding input

$0.088per 1M tokens

Context

128K

Max output

—

Capability

—

First token

—

30-day uptime

—

Faster, lower-cost Embed 5 model sharing the Pro embedding space, with 128K context and up to 2048-dim vectors.

Embedding input

$0.132per 1M tokens

Context

128K

Max output

—

Capability

—

First token

—

30-day uptime

—

Cohere's highest-quality Embed 5 model for multilingual enterprise retrieval, with 128K context and up to 2048-dim vectors.

Input

from $1.80list $2.00

Output

from $9.00list $10.00

Discount

-10%

Unit

per 1M tokens

Cache read

from $0.09list $0.10

Context

1.1M

Max output

128K

Capability

168

First token

—

30-day uptime

—

Providers

lowest of 2 providers

OpenAI's upgraded GPT-6 Sol with near-Astra performance for complex coding, computer use, and professional work.

Input

from $2.20

Output

from $11.00

Unit

per 1M tokens

Cache read

from $0.11

Context

1M

Max output

128K

Capability

167

First token

—

30-day uptime

—

Providers

lowest of 3 providers

Anthropic's Claude Sonnet 5.5 for everyday coding, agentic tasks, and polished documents, faster than Sonnet 5 at the same price.

Input

from $0.1606

Output

from $0.6424

Unit

per 1M tokens

Cache read

from $0.01606

Context

1M

Max output

524K

Capability

—

First token

—

30-day uptime

—

Providers

lowest of 3 providers

MiniMax's frontier multimodal coding model with a 1M context window and tunable thinking depth.

Input

from $4.40

Output

from $22.00

Unit

per 1M tokens

Cache read

from $0.22

Context

1M

Max output

128K

Capability

168

First token

—

30-day uptime

—

Providers

lowest of 3 providers

Anthropic's Claude Opus 5.5 for long-running agentic coding and knowledge work, priced below Opus 5 with adaptive thinking always on.

Input

$3.00

Output

$15.00

Unit

per 1M tokens

Cache read

$0.30

Context

1M

Max output

131K

Capability

—

First token

—

30-day uptime

—

Fireworks' specialized model built on Kimi K3 with shorter reasoning traces for efficient agentic workloads.

Input

$0.45list $0.50

Output

$8.10list $9.00

Discount

-10%

Unit

per 1M tokens

Cache read

$0.45list $0.50

Context

8K

Max output

16K

Capability

—

First token

—

30-day uptime

—

Providers

lowest of 2 providers

Google's flagship text-to-speech model for studio-grade voice fidelity, expressive acting, regional accents, and stable long-form multi-speaker audio.

Input

$0.45list $0.50

Output

$5.40list $6.00

Discount

-10%

Unit

per 1M tokens

Cache read

$0.45list $0.50

Context

8K

Max output

16K

Capability

—

First token

—

30-day uptime

—

Providers

lowest of 2 providers

Google's fast, cost-efficient text-to-speech model for high-volume and real-time single- and multi-speaker speech.

Input

from $0.09list $0.10

Output

from $0.45list $0.50

Discount

-10%

Unit

per 1M tokens

Cache read

from $0.009list $0.01

Context

1.1M

Max output

128K

Capability

157

First token

—

30-day uptime

—

Providers

lowest of 2 providers

OpenAI's cost-efficient GPT-6 model for high-volume workloads such as extraction, summarization, and routing.

Input

from $1.80list $2.00

Output

from $9.00list $10.00

Discount

-10%

Unit

per 1M tokens

Cache read

from $0.18list $0.20

Context

1.1M

Max output

128K

Capability

164

First token

—

30-day uptime

—

Providers

lowest of 2 providers

OpenAI's frontier GPT-6 model for enterprise agents, coding, writing, and complex knowledge work.

Input

$0.1606

Output

$0.3212

Unit

per 1M tokens

Cache read

$0.003212

Context

1M

Max output

131K

Capability

—

First token

—

30-day uptime

—

Xiaomi's efficient omni-modal reasoning model (309B total, 15B active) for high-volume coding and agent workflows.

Input

$0.4818

Output

$0.9636

Unit

per 1M tokens

Cache read

$0.004016

Context

1M

Max output

131K

Capability

—

First token

—

30-day uptime

—

Xiaomi's flagship omni-modal reasoning model (1.02T total, 42B active) for long-horizon agents, coding, and research.

Input

$1.1242

Output

$3.212

Unit

per 1M tokens

Cache read

$0.056212

Context

1M

Max output

66K

Capability

—

First token

—

30-day uptime

—

StepFun's flagship 600B MoE model (27B active) for agentic coding and professional knowledge work, with 1M context and vision input.

Input

$0.3212

Output

$1.1242

Unit

per 1M tokens

Cache read

$0.091544

Context

1M

Max output

131K

Capability

—

First token

—

30-day uptime

—

Providers

lowest of 2 providers

High-speed serving option of GLM-5.3-Flash for coding and agent workflows.

Input

$0.113

Output

$0.382

Unit

per 1M tokens

Cache read

$0.014

Context

1M

Max output

131K

Capability

—

First token

—

30-day uptime

—

Qwen3.8 omni-modal model for audio and video understanding with text output.

Input

$0.1168

Output

$0.3942

Unit

per 1M tokens

Cache read

$0.02336

Context

1M

Max output

256K

Capability

—

First token

—

30-day uptime

—

A lower-cost Doubao Seed 2.1 variant for everyday multimodal workloads.

Input

$0.876

Output

$4.38

Unit

per 1M tokens

Cache read

$0.1752

Context

1M

Max output

256K

Capability

—

First token

—

30-day uptime

—

ByteDance's flagship Doubao Seed 2.1 model with 1M context for coding and agent workloads.

Input

$0.675list $0.75

Output

$4.05list $4.50

Discount

-10%

Unit

per 1M tokens

Cache read

$0.675list $0.75

Context

131K

Max output

66K

Capability

—

First token

—

30-day uptime

—

Google Gemini 3.8 real-time bidirectional Live API model for low-latency voice agents. Supports text, image, and audio input with text and audio output.

Input

$0.675list $0.75

Output

$4.05list $4.50

Discount

-10%

Unit

per 1M tokens

Cache read

$0.675list $0.75

Context

131K

Max output

66K

Capability

—

First token

—

30-day uptime

—

Google Gemini 3.8 Live with background reasoning for real-time voice agents that need deeper thinking. Supports text, image, and audio input with text and audio output.

Input

$0.042per 1M tokens

Context

64K

Max output

—

Capability

—

First token

—

30-day uptime

—

TypeSafe System One decision model. Evaluates a state against typed questions and returns typed answers (noul probability / choice / score / box) with calibrated confidence; it does not generate text. Image assets (and box questions) require the TypeSafe-Danger-Zone-Images: true header.

Input

$1.0439

Output

$4.3362

Unit

per 1M tokens

Cache read

$0.27302

Context

1M

Max output

1M

Capability

—

First token

—

30-day uptime

—

Moonshot AI's K2.8 preview for coding and agent work, close to Kimi K3 with more efficient thinking and 1M context.

Input

from $0.073list $0.146-50%

Output

from $0.292list $0.584-50%

Unit

per 1M tokens

Cache read

from $0.00292

Context

1M

Max output

384K

Capability

156

First token

—

30-day uptime

—

Providers

lowest of 4 providers

DeepSeek V4.1 Flash: natively multimodal vision understanding with 1M context and 384K max output.

Image output

$27.00list $30.00per 1M tokens

Context

128K

Max output

—

Capability

—

First token

—

30-day uptime

—

Providers

lowest of 2 providers

OpenAI's fastest image model for high-quality everyday generation and editing, with text and image inputs.

Image output

$27.00list $30.00per 1M tokens

Context

128K

Max output

—

Capability

—

First token

—

30-day uptime

—

Providers

lowest of 2 providers

OpenAI's most capable image model for generation and editing where editing precision matters most, with text and image inputs.

Input

from $9.00list $10.00

Output

from $45.00list $50.00

Discount

-10%

Unit

per 1M tokens

Cache read

from $0.90list $1.00

Context

1.1M

Max output

128K

Capability

169

First token

—

30-day uptime

—

Providers

lowest of 2 providers

OpenAI's most capable model, built for the hardest end-to-end work: complex reasoning, coding, computer use, research, and document creation.

Input

$0.675list $0.75

Output

$3.375list $3.75

Discount

-10%

Unit

per 1M tokens

Cache read

$0.0675list $0.075

Context

1M

Max output

64K

Capability

158

First token

—

30-day uptime

—

Providers

lowest of 2 providers

Google's workhorse model for coding and agents, with sharper multi-turn coding, rapid refactoring, code synthesis, and tighter tool-execution logic.

Input

$1.65

Output

$4.951

Unit

per 1M tokens

Cache read

$0.206

Context

1M

Max output

131K

Capability

—

First token

—

30-day uptime

—

Upgraded Qwen3.8 Max snapshot with stronger coding, multi-tool agent work, and vision understanding.

Input

from $11.00

Output

from $55.00

Unit

per 1M tokens

Cache read

from $0.275

Context

1M

Max output

128K

Capability

166

First token

—

30-day uptime

—

Providers

lowest of 2 providers

Claude Fable 5.1 extends Claude Fable 5 with stronger long-running agentic coding, knowledge work, and research. Same $10/$50 rates with cache reads at a quarter of the cost.

Input

$0.876

Output

$2.628

Unit

per 1M tokens

Cache read

$0.0438

Context

1M

Max output

64K

Capability

—

First token

—

30-day uptime

—

Tencent Hy4 preview reasoning model served by TokenHub for coding agents, complex tool calls, and multi-step agentic workflows.

Input

$1.80list $2.00

Output

$10.80list $12.00

Discount

-10%

Unit

per 1M tokens

Cache read

$1.80list $2.00

Context

131K

Max output

66K

Capability

—

First token

—

30-day uptime

—

Providers

lowest of 2 providers

Google Gemini 3.5 speech-to-text for audio files up to one hour, with language detection, speaker diarization, word-level timestamps, and custom vocabulary.

Input

$3.15list $3.50

Output

$18.90list $21.00

Discount

-10%

Unit

per 1M tokens

Cache read

$3.15list $3.50

Context

131K

Max output

66K

Capability

—

First token

—

30-day uptime

—

Google Gemini 3.5 low-latency speech-to-text over the Live API for bidirectional streaming audio transcription.

Input

from $0.0075list $0.15

Output

from $0.025list $0.50

Discount

-95%

Unit

per 1M tokens

Cache read

from $0.0015list $0.03

Context

1M

Max output

131K

Capability

153

First token

—

30-day uptime

—

Providers

lowest of 6 providers

Zhipu's first natively multimodal GLM-5 model (320B/18B MoE) with 1M context and visual coding capabilities.

Video

$0.022per second

Context

4K

Max output

—

Capability

—

First token

—

30-day uptime

—

xAI's lowest-cost video model for simple text-to-video clips with audio.

Input

$0.113

Output

$0.382

Unit

per 1M tokens

Cache read

$0.014

Context

1M

Max output

131K

Capability

—

First token

—

30-day uptime

—

Fast, low-cost Qwen3.8 multimodal model with 1M context for coding, agents, and visual understanding.

Input

from $0.07list $1.40

Output

from $0.22list $4.40

Discount

-95%

Unit

per 1M tokens

Cache read

from $0.013list $0.26

Context

1M

Max output

131K

Capability

156

First token

—

30-day uptime

—

Providers

lowest of 7 providers

Z.ai's latest flagship reasoning model with a 1M-token context window and controllable thinking length.

Input

from $0.495

Output

from $1.639

Unit

per 1M tokens

Cache read

from $0.055

Context

262K

Max output

131K

Capability

—

First token

—

30-day uptime

—

Providers

lowest of 3 providers

Qwen3.8 27B open-weight dense vision-language model with default-on thinking for coding and agentic tasks.

Input

$0.675list $0.75

Output

$3.375list $3.75

Discount

-10%

Unit

per 1M tokens

Cache read

$0.0675list $0.075

Context

1M

Max output

64K

Capability

158

First token

—

30-day uptime

—

Google's most intelligent workhorse model for coding and agentic workflows.

Input

$0.1104list $0.138

Output

$0.22list $0.275

Discount

-20%

Unit

per 1M tokens

Cache read

$0.0224list $0.028

Context

1M

Max output

384K

Capability

155

First token

—

30-day uptime

—

DeepSeek V4 Flash 0731 weight snapshot with 1M context and 384K max output.

Transcription

$0.000068list $0.000076-11%per second

Context

128K

Max output

—

Capability

—

First token

—

30-day uptime

—