opencode-kimi-full

An opencode plugin for the Kimi For Coding plan.

This plugin authenticates the same way the official kimi-cli does and mirrors its wire shape, so opencode's requests to Moonshot's /coding endpoint match kimi-cli byte-for-byte, including OAuth, fingerprint headers, session-scoped prompt caching, and paired thinking / reasoning-effort fields for higher fidelity. The Kimi specific extras exposed by the /coding endpoint used by kimi-cli are not implemented in opencode. Bonus side effect; users with access to Kimi K2.6 Code Preview will get access to it in opencode with this plugin.

Contributor and agent documentation lives in AGENTS.md.

---

Requirements

• opencode ≥ 1.4.6
• A Kimi account with an active Kimi For Coding subscription (the same plan that works with kimi-cli)

Install

cd ~/.opencode
bun add opencode-kimi-full

From a local checkout

cd ~/.opencode
bun add /path/to/opencode-kimi-full

Configure

Add the plugin and a provider entry to opencode.json (or ~/.config/opencode/opencode.json):

{
  "$schema": "https://opencode.ai/config.json",
  "plugin": ["opencode-kimi-full"],
  "provider": {
    "kimi-for-coding-oauth": {
      "name": "Kimi For Coding (OAuth)",
      "npm": "@ai-sdk/openai-compatible",
      "options": {
        "baseURL": "https://api.kimi.com/coding/v1"
      },
      "models": {
        "kimi-for-coding": {
          "name": "Kimi For Coding",
          "reasoning": true,
          "options": {},
          "variants": {
            "off":    { "reasoning_effort": "off" },
            "auto":   { "reasoning_effort": "auto" },
            "low":    { "reasoning_effort": "low" },
            "medium": { "reasoning_effort": "medium" },
            "high":   { "reasoning_effort": "high" }
          }
        }
      }
    }
  }
}

Two identifiers are load-bearing:

• provider id kimi-for-coding-oauth — the plugin's auth and chat.params hooks match on it.
• model id kimi-for-coding — a stable opencode-side alias. The plugin rewrites the wire model field to whatever /coding/v1/models reports for your account (e.g. kimi-for-coding on K2.6 tiers, k2p5 on K2.5 tiers). Both tiers use identical config.

Note. The provider id is intentionally not kimi-for-coding. That id is already published by models.dev and points at a static-API-key flow that routes to K2.5. Using a distinct id keeps the two paths from colliding under a single opencode auth login entry.

Log in

opencode auth login -p kimi-for-coding-oauth

The plugin returns a verification URL and user code. After browser approval it polls the device-auth endpoint, queries /coding/v1/models to discover the model id and context length your account is entitled to, prints a ready-to-paste config block with those values filled in, and persists everything (tokens + discovered metadata) through opencode's auth.json. The model list is re-checked on every token refresh, mirroring kimi-cli's refresh_managed_models. Access tokens have a ~15 minute TTL and refresh automatically; refresh tokens last ~30 days.

Use

Select kimi-for-coding-oauth/kimi-for-coding in opencode.

Press Ctrl+T to pick a reasoning variant (off, auto, low, medium, high). off disables thinking; auto lets Moonshot pick dynamically; low/medium/high pin the effort level.

---

Why this plugin exists

There are two ways to talk to Moonshot's Kimi For Coding plan today: the way kimi-cli does it, and the way opencode does it. They target different endpoints and use different authentication. This plugin brings kimi-cli parity into opencode.

How kimi-cli does it. OAuth device-code flow against auth.moonshot.cn with scope: kimi-code, producing a short-lived JWT. Requests go to https://api.kimi.com/coding/v1 (OpenAI-compatible) with the JWT as the bearer token, seven X-Msh-* fingerprint headers, a stable ~/.kimi/device_id, and per-request extras: prompt_cache_key (an opt-in, session-scoped cache key) and paired thinking.type + reasoning_effort. The backend routes this token to K2.6.

How opencode does it out of the box. opencode auth login selects a Kimi For Coding provider from the catalog and prompts for a KIMI_API_KEY (a static sk-kimi-... key). The catalog entry uses @ai-sdk/anthropic against api.kimi.com/coding, which is valid since the endpoint exposes both OpenAI-compatible and Anthropic-compatible routes. No Kimi-specific request extras are sent. The backend currently routes a static sk-kimi-... key to K2.5.

What this plugin gives you. Everything kimi-cli does, inside opencode:

• OAuth device flow with scope: kimi-code, so you land on K2.6 (if you have access).
• prompt_cache_key set to opencode's session id, for session-scoped cache reuse.
• Paired thinking + reasoning_effort fields.
• The seven X-Msh-* headers and a kimi-cli-shaped User-Agent.
• ~/.kimi/device_id shared with a locally-installed kimi-cli.
• Tokens stored in opencode's auth.json under a dedicated provider id, so the plugin and kimi-cli keep independent refresh-token chains and do not invalidate each other.
• Streaming, reasoning_content deltas, and tool-call schemas are handled upstream by @ai-sdk/openai-compatible — not reimplemented here.

Relationship to potential upstream fixes

Two upstream changes would narrow the gap, but neither would make this plugin redundant:

• If Moonshot routes sk-kimi-... keys to K2.6, opencode's built-in path reaches K2.6 too, but still without prompt_cache_key or the paired reasoning fields. Explicit session-scoped cache reuse stays unavailable on that path (any automatic prefix caching Moonshot may do is orthogonal and would apply to both paths), and reasoning is controlled on the Anthropic route via thinking.budget_tokens — the paired reasoning_effort: low|medium|high knob that kimi-cli exposes has no equivalent there.
• If opencode ships a native Kimi For Coding OAuth, the auth story converges, but the request-field gap stays until opencode's provider code emits these exact fields for /coding. kimi-cli is Moonshot's first-party client and targets the OpenAI-compatible route, so mirroring its wire shape is the lowest-risk way to stay aligned with upstream. Fingerprint parity (same X-Msh-Device-Id and headers, kimi-cli-shaped UA) and independent refresh-token chains are unlikely to be replicated by a first-party integration.

Request fields in detail

Field	Wire shape	Purpose
prompt_cache_key	top-level body, snake_case, set to opencode's sessionID	Opt-in, session-scoped cache key, mirroring kimi-cli.
thinking + reasoning_effort	thinking: { type: "enabled" | "disabled" } with sibling reasoning_effort: "low" | "medium" | "high"	Sent together, matching kimi-cli.
Seven X-Msh-* headers + UA	User-Agent, X-Msh-Platform, X-Msh-Version, X-Msh-Device-Name, X-Msh-Device-Model, X-Msh-Device-Id, X-Msh-OS-Version	Matches kimi-cli's _kimi_default_headers() at the pinned KIMI_CLI_VERSION.
~/.kimi/device_id	UUID persisted on disk, embedded in X-Msh-Device-Id	Sends the same X-Msh-Device-Id as a locally-installed kimi-cli.

Effort-to-field mapping, taken verbatim from kimi-cli:

user effort	reasoning_effort	thinking
auto	(omitted)	(omitted) — server picks dynamically
off	(omitted)	{ type: "disabled" }
low / medium / high	same string	{ type: "enabled" }

Files the plugin touches

Path	Purpose
~/.kimi/device_id	Stable UUID used in X-Msh-Device-Id. Shared with kimi-cli.
<opencode data>/auth.json	Token storage, managed by opencode through client.auth.*.

No other state is persisted. Credentials are never written to ~/.kimi/credentials/; that path belongs to kimi-cli, and sharing it would cause refresh-token races between the two clients.

Architecture at a glance

┌────────────── opencode core ─────────────┐
│                                          │
│  auth.login ─▶ plugin.auth.authorize()   │  device-code flow, poll
│                 └─▶ oauth.ts             │
│                                          │
│  chat ──────▶ plugin.loader()            │  custom fetch that:
│                 ├─▶ ensureFresh()        │   • proactive refresh
│                 └─▶ kimiHeaders()        │   • 7 X-Msh-* headers
│                                          │   • 401 → force-refresh + retry
│  chat.params ─▶ plugin "chat.params"     │  thinking / reasoning_effort /
│                                          │  prompt_cache_key
└──────────────────────────────────────────┘

A full description of the invariants that keep this working is in AGENTS.md, under "Architecture" and "Contracts to keep intact".

License

MIT.