@@ -0,0 +1,94 @@
1+---
2+summary: "Cerebras setup (auth + model selection)"
3+title: "Cerebras"
4+read_when:
5+ - You want to use Cerebras with OpenClaw
6+ - You need the Cerebras API key env var or CLI auth choice
7+---
8+9+[Cerebras](https://www.cerebras.ai) provides high-speed OpenAI-compatible inference.
10+11+| Property | Value |
12+| -------- | ---------------------------- |
13+| Provider | `cerebras` |
14+| Auth | `CEREBRAS_API_KEY` |
15+| API | OpenAI-compatible |
16+| Base URL | `https://api.cerebras.ai/v1` |
17+18+## Getting Started
19+20+<Steps>
21+<Step title="Get an API key">
22+Create an API key in the [Cerebras Cloud Console](https://cloud.cerebras.ai).
23+</Step>
24+<Step title="Run onboarding">
25+```bash
26+openclaw onboard --auth-choice cerebras-api-key
27+```
28+</Step>
29+<Step title="Verify models are available">
30+```bash
31+openclaw models list --provider cerebras
32+```
33+</Step>
34+</Steps>
35+36+### Non-Interactive Setup
37+38+```bash
39+openclaw onboard --non-interactive \
40+ --mode local \
41+ --auth-choice cerebras-api-key \
42+ --cerebras-api-key "$CEREBRAS_API_KEY"
43+```
44+45+## Built-In Catalog
46+47+OpenClaw ships a static Cerebras catalog for the public OpenAI-compatible endpoint:
48+49+| Model ref | Name | Notes |
50+| ----------------------------------------- | -------------------- | -------------------------------------- |
51+| `cerebras/zai-glm-4.7` | Z.ai GLM 4.7 | Default model; preview reasoning model |
52+| `cerebras/gpt-oss-120b` | GPT OSS 120B | Production reasoning model |
53+| `cerebras/qwen-3-235b-a22b-instruct-2507` | Qwen 3 235B Instruct | Preview non-reasoning model |
54+| `cerebras/llama3.1-8b` | Llama 3.1 8B | Production speed-focused model |
55+56+<Warning>
57+Cerebras marks `zai-glm-4.7` and `qwen-3-235b-a22b-instruct-2507` as preview models, and `llama3.1-8b` / `qwen-3-235b-a22b-instruct-2507` are documented for deprecation on May 27, 2026. Check Cerebras' supported-models page before relying on them for production.
58+</Warning>
59+60+## Manual Config
61+62+The bundled plugin usually means you only need the API key. Use explicit
63+`models.providers.cerebras` config when you want to override model metadata:
64+65+```json5
66+{
67+ env: { CEREBRAS_API_KEY: "sk-..." },
68+ agents: {
69+ defaults: {
70+ model: { primary: "cerebras/zai-glm-4.7" },
71+ },
72+ },
73+ models: {
74+ mode: "merge",
75+ providers: {
76+ cerebras: {
77+ baseUrl: "https://api.cerebras.ai/v1",
78+ apiKey: "${CEREBRAS_API_KEY}",
79+ api: "openai-completions",
80+ models: [
81+ { id: "zai-glm-4.7", name: "Z.ai GLM 4.7" },
82+ { id: "gpt-oss-120b", name: "GPT OSS 120B" },
83+ ],
84+ },
85+ },
86+ },
87+}
88+```
89+90+<Note>
91+If the Gateway runs as a daemon (launchd/systemd), make sure `CEREBRAS_API_KEY`
92+is available to that process, for example in `~/.openclaw/.env` or through
93+`env.shellEnv`.
94+</Note>
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。