Models

Me Write Code runs against any model your provider exposes. The defaults are chosen per provider and re-evaluated on each release based on the proof-bench eval harness.

Default models per provider

Provider	Default model
Anthropic	claude-opus-4-6
OpenAI	gpt-5.4
Google	gemini-2.5-pro
OpenRouter	openai/gpt-5.1-codex
Groq	openai/gpt-oss-120b
Cerebras	zai-glm-4.7

Override per session:

mewrite --model claude-opus-4-7
mewrite --model openai/gpt-5
mewrite --model claude-sonnet-4:high   # thinking level high

Inside the TUI, /model opens the picker.

Thinking levels

Models that support extended thinking accept a suffix:

Level	Use case
:off	Fastest, lowest cost. Default.
:minimal	Light reasoning. Routine edits.
:low	Default thinking budget for most providers.
:medium	Multi-file refactors.
:high	Cross-cutting concerns, architectural changes.
:xhigh	Hard debugging, complex algorithms.

Cycle in TUI with Shift+Tab.

Architect / editor split

Use a strong model to plan, a cheaper model to execute. Drops cost ~3-5× on long sessions.

/architect set architectModel=claude-opus-4-7 editorModel=claude-haiku-4

Or in ~/.mewrite/agent/settings.json:

{
    "model": "claude-sonnet-4",
    "modes": {
        "architect": { "model": "claude-opus-4-7" },
        "editor": { "model": "claude-haiku-4" }
    }
}

Per-subagent models

Each subagent declares its own model:

---
description: "Run unit tests and report failures"
model: "claude-haiku-4"   # cheap, since it just shells out
tools: [Bash, Read]
---

Subagent results are summarized to ≤500 tokens before re-entering the parent's context — letting you spend on Haiku instead of Opus for repetitive subtasks.

Capability tiers

Instead of a concrete model id, a subagent can request a capability tier — tier:fast, tier:normal, or tier:strong:

---
description: "Run unit tests and report failures"
model: "tier:fast"   # cheap model of whatever provider you're on
tools: [Bash, Read]
---

A tier resolves to a curated model within your current provider, so the same agent definition works whether you're authed to Anthropic, OpenAI, Google, or Bedrock — no provider-specific id to hardcode.

• tier:fast — cheap/small model. High-volume, repeatable, small tasks.
• tier:normal — the everyday model. Editing, testing, implementation.
• tier:strong — the top model. Orchestration and planning.

If a tier has no curated model for your provider (e.g. a router or a custom provider), the subagent falls back to the provider's normal, then to the parent's model with a warning. Tiers are best-effort and not pinned to an exact model — the curated mapping can change between releases.

Set a default tier for all subagents that don't pin a model via the subagentModel setting in settings.json:

{ "subagentModel": "tier:fast" }

Model registry

Provider/model definitions ship with Me Write Code and can be refreshed without upgrading the CLI:

mewrite models update

Override per-machine in ~/.mewrite/agent/registry.json. See Provider Registry for schema.

Cost-aware defaults

Me Write Code's first-run wizard suggests Haiku/Flash for the default model on free OAuth accounts to avoid surprise bills. Upgrade with /model once you've validated the workflow.