Gyro

Published catalog

Model Catalog

Providers and models Gyro ships.

Catalog revision 2026-11-01.2

OpenAI

9 models

  1. GPT-6.1 Sol

    gpt-6.1-sol

    Codex CLI. Near-Astra performance for coding and professional work.

    Context
    500,000 tokens
    Reasoning
    • low default
    • medium
    • high
    • xhigh
    • max
    • ultra

  2. GPT-6.1 Terra

    gpt-6.1-terra

    Codex CLI. GPT-6.1 Terra Codex model.

    Context
    500,000 tokens
    Reasoning
    • low default
    • medium
    • high
    • xhigh
    • max
    • ultra

  3. GPT-6 Astra

    gpt-6-astra

    Complex reasoning, coding, and long-running work.

    Context
    272,000 tokens
    Reasoning
    • low
    • medium default
    • high
    • xhigh
    • max
    • ultra

  4. GPT-6 Sol

    gpt-6-sol

    Codex CLI. Workhorse model for coding and everyday work.

    Context
    272,000 tokens
    Reasoning
    • low
    • medium default
    • high
    • xhigh
    • max
    • ultra

  5. GPT-6 Luna

    gpt-6-luna

    Codex CLI. Fast and affordable model for easier tasks.

    Context
    272,000 tokens
    Reasoning
    • low
    • medium default
    • high
    • xhigh
    • max

  6. GPT-5.6 Sol

    gpt-5.6-sol

    Codex CLI. Frontier model for complex professional work.

    Context
    272,000 tokens
    Reasoning
    • low default
    • medium
    • high
    • xhigh
    • max
    • ultra

  7. GPT-5.6 Terra

    gpt-5.6-terra

    Balances intelligence and cost for everyday work.

    Context
    272,000 tokens
    Reasoning
    • low
    • medium default
    • high
    • xhigh
    • max
    • ultra

  8. GPT-5.6 Luna

    gpt-5.6-luna

    Codex CLI. Fast, cost-sensitive model for lighter workloads.

    Context
    272,000 tokens
    Reasoning
    • low
    • medium default
    • high
    • xhigh
    • max

  9. GPT-5.5

    gpt-5.5

    Flagship model for complex reasoning and coding.

    Context
    272,000 tokens
    Reasoning
    • low
    • medium default
    • high
    • xhigh

Anthropic

9 models

  1. Claude Fable 5.1

    claude-fable-5-1

    Demanding reasoning and long-running agentic work.

    Context
    1,000,000 tokens
    Reasoning
    • low
    • medium
    • high default
    • xhigh
    • max

  2. Claude Fable 5

    claude-fable-5

    Previous-generation Fable model for complex work.

    Context
    1,000,000 tokens
    Reasoning
    • low
    • medium
    • high default
    • xhigh
    • max

  3. Claude Opus 5.5

    claude-opus-5-5

    Claude Code. Long-running agentic coding and knowledge work.

    Context
    200,000 tokens
    Reasoning
    • low
    • medium default
    • high
    • xhigh
    • max

  4. Claude Opus 5

    claude-opus-5

    Frontier model for complex agentic coding and long-horizon work.

    Context
    1,000,000 tokens
    Reasoning
    • low
    • medium
    • high default
    • xhigh
    • max

  5. Claude Opus 4.8

    claude-opus-4-8

    Strong model for complex agentic coding.

    Context
    1,000,000 tokens
    Reasoning
    • low
    • medium
    • high default
    • xhigh
    • max

  6. Claude Sonnet 5.5

    claude-sonnet-5-5

    Claude Code. Best speed and intelligence balance.

    Context
    1,000,000 tokens
    Reasoning
    • low
    • medium default
    • high
    • xhigh
    • max

  7. Claude Sonnet 5

    claude-sonnet-5

    Best speed and intelligence balance.

    Context
    1,000,000 tokens
    Reasoning
    • low
    • medium
    • high default
    • xhigh
    • max

  8. Claude Haiku 5.5

    claude-haiku-5-5

    Claude Code. Fast model for lighter coding and high-volume tasks.

    Context
    1,000,000 tokens
    Reasoning
    • low
    • medium default
    • high
    • xhigh
    • max

  9. Claude Haiku 4.5

    claude-haiku-4-5

    Claude Code. Fastest option for lighter work.

    Context
    200,000 tokens

Kimi

3 models

  1. Kimi K3.1 Pro

    k3.1-pro

    Kimi Code. Flagship of the K3.1 series. Up to 1M context on Pro and Allegretto; Plus and Moderato stay at 262,144. Thinking is required.

    Context
    262,144 tokens
    Reasoning
    • low
    • high
    • max default

  2. Kimi K3.1 Pro Highspeed

    k3.1-pro-highspeed

    Kimi Code. The same K3.1 Pro model, about 2-3x faster. Pro and Allegretto only, with a 1,048,576 context window.

    Context
    1,048,576 tokens
    Reasoning
    • low
    • high
    • max default

  3. Kimi K3

    k3

    Flagship coding model. Up to 1M context on eligible Kimi Code plans.

    Context
    262,144 tokens
    Reasoning
    • low
    • high default
    • max

xAI

5 models

  1. Grok 4.7

    grok-4.7

    xAI's frontier model for coding, agentic tasks, and knowledge work.

    Context
    500,000 tokens
    Reasoning
    • low
    • medium
    • high default
    • xhigh

  2. Grok 4.7 Fast

    grok-4.7-build-fast

    The same Grok 4.7 model on faster Grok Build infrastructure, at higher usage.

    Context
    500,000 tokens
    Reasoning
    • low
    • medium
    • high default
    • xhigh

  3. Grok 4.6

    grok-4.6

    Previous xAI frontier model for coding, agentic tasks, and knowledge work.

    Context
    500,000 tokens
    Reasoning
    • low
    • medium
    • high default
    • xhigh

  4. Grok 4.5

    grok-4.5

    xAI's coding-capable model through Grok Build.

    Context
    500,000 tokens
    Reasoning
    • low
    • medium
    • high default

  5. Grok 4.3

    grok-4.3

    General xAI model for chat and reasoning.

    Context
    1,000,000 tokens
    Reasoning
    • low default
    • medium
    • high
    • xhigh

Google Antigravity

1 model

  1. Automatic

    antigravity-default

    Uses the model configured in Google Antigravity. Connect to discover available models.

Cursor

1 model

  1. Cursor default

    cursor-default

    Uses the model configured in your local Cursor CLI.

OpenCode

1 model

  1. OpenCode default

    opencode-default

    Uses the model configured in your local OpenCode CLI.

DeepSeek

2 models

  1. DeepSeek V4.1 Flash

    deepseek-flash

    Fast general coding model, and the cheaper of the two.

    Context
    1,000,000 tokens
    Reasoning
    • low
    • high default
    • max

  2. DeepSeek V4 Pro

    deepseek-v4-pro

    Deeper reasoning for harder problems, at a higher price.

    Context
    1,000,000 tokens
    Reasoning
    • low
    • high default
    • max

Mistral

4 models

  1. Mistral Medium 3.5

    mistral-medium-latest

    Frontier-class model tuned for agentic and coding work.

    Context
    256,000 tokens

  2. Mistral Large 3

    mistral-large-latest

    Open-weight flagship for general multimodal work.

    Context
    256,000 tokens

  3. Mistral Small 4

    mistral-small-latest

    Cheapest of the three, and the fastest to first token.

    Context
    256,000 tokens

  4. Codestral

    codestral-latest

    Tuned for code completion and editing.

    Context
    128,000 tokens

OpenRouter

12 models

  1. Claude Sonnet 5

    anthropic/claude-sonnet-5

    OpenRouter API. Strong general coding model. Claude Code also serves 1M on its own connection.

    Context
    1,000,000 tokens
    Reasoning
    • low
    • medium
    • high default
    • xhigh
    • max

  2. Claude Opus 5

    anthropic/claude-opus-5

    OpenRouter API. Anthropic's frontier model. Claude Code also serves 1M on its own connection.

    Context
    1,000,000 tokens
    Reasoning
    • low
    • medium
    • high default
    • xhigh
    • max

  3. GPT-6 Astra

    openai/gpt-6-astra

    OpenRouter API. OpenAI's frontier model. Codex CLI serves 272K on its own connection.

    Context
    1,050,000 tokens
    Reasoning
    • low
    • medium default
    • high
    • xhigh
    • max

  4. Gemini 3.8 Flash

    google/gemini-3.8-flash

    OpenRouter API. Long-context reasoning.

    Context
    1,048,576 tokens
    Reasoning
    • low
    • medium default
    • high

  5. Grok 4.7

    x-ai/grok-4.7

    OpenRouter API. Serves 500K, separate from the Grok Build CLI's 256K.

    Context
    500,000 tokens
    Reasoning
    • low
    • medium
    • high default
    • xhigh

  6. Grok 4.6

    x-ai/grok-4.6

    OpenRouter API. Serves 500K, separate from the Grok Build CLI's 256K.

    Context
    500,000 tokens
    Reasoning
    • low
    • medium
    • high default
    • xhigh

  7. DeepSeek V4.1 Flash

    deepseek/deepseek-v4.1-flash

    OpenRouter API. Low-cost general model. DeepSeek's own API serves 1M.

    Context
    1,048,576 tokens
    Reasoning
    • low
    • high default
    • max

  8. Kimi K3.1 Pro

    moonshotai/kimi-k3.1-pro

    OpenRouter API. Flagship open-weight model with a 1,048,576 context. Kimi Code's offline baseline for this model is 262,144.

    Context
    1,048,576 tokens
    Reasoning
    • low
    • high
    • max default

  9. Kimi K3

    moonshotai/kimi-k3

    OpenRouter API. Open-weights agentic model. Kimi Code's offline baseline is 262,144.

    Context
    1,048,576 tokens
    Reasoning
    • low
    • high
    • max default

  10. GLM 5.3

    z-ai/glm-5.3

    OpenRouter API. Open-weights coding model.

    Context
    1,048,576 tokens
    Reasoning
    • low
    • high
    • max default

  11. Qwen3.8 Max

    qwen/qwen3.8-max-0902

    OpenRouter API. Qwen's frontier model.

    Context
    1,000,000 tokens
    Reasoning
    • low
    • medium
    • high
    • xhigh default

  12. Llama 3.3 70B

    meta-llama/llama-3.3-70b-instruct

    Open-weights general model, routed by OpenRouter.

    Context
    131,072 tokens