romtuck/openclaw
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: add Nebius Token Factory provider

Browse Source
This commit is contained in:
kiranchilledout 2026-01-30 07:04:51 +05:30
parent 4583f88626
commit 84a1994890
10 changed files with 443 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

40
docs/concepts/model-providers.md
View File

@ -239,6 +239,46 @@ MiniMax is configured via `models.providers` because it uses custom endpoints:
See [/providers/minimax](/providers/minimax) for setup details, model options, and config snippets.
### Nebius Token Factory
Nebius provides OpenAI-compatible inference for frontier and open models:
- Provider: `nebius`
- Auth: `NEBIUS_API_KEY`
- Example model: `nebius/Qwen/Qwen3-32B-fast`
- CLI: `moltbot onboard --auth-choice nebius-api-key`
```json5
{
agents: {
defaults: { model: { primary: "nebius/Qwen/Qwen3-32B-fast" } }
},
models: {
mode: "merge",
providers: {
nebius: {
baseUrl: "https://api.tokenfactory.nebius.com/v1",
apiKey: "${NEBIUS_API_KEY}",
api: "openai-completions",
models: [
{
id: "Qwen/Qwen3-32B-fast",
name: "Qwen3 32B Fast",
reasoning: false,
input: ["text"],
cost: { input: 0, output: 0, cacheRead: 0, cacheWrite: 0 },
contextWindow: 128000,
maxTokens: 8192
}
]
}
}
}
}
```
See [/providers/nebius](/providers/nebius) for setup details and available models.
### Ollama
Ollama is a local LLM runtime that provides an OpenAI-compatible API:

42
docs/gateway/configuration.md
View File

@ -2472,6 +2472,48 @@ Notes:
- Set `KIMICODE_API_KEY` in the environment or use `moltbot onboard --auth-choice kimi-code-api-key`.
- Model ref: `kimi-code/kimi-for-coding`.
### Nebius Token Factory
Use Nebius's OpenAI-compatible endpoint:
```json5
{
env: { NEBIUS_API_KEY: "sk-..." },
agents: {
defaults: {
model: { primary: "nebius/Qwen/Qwen3-32B-fast" },
models: { "nebius/Qwen/Qwen3-32B-fast": { alias: "Qwen3 32B" } }
}
},
models: {
mode: "merge",
providers: {
nebius: {
baseUrl: "https://api.tokenfactory.nebius.com/v1",
apiKey: "${NEBIUS_API_KEY}",
api: "openai-completions",
models: [
{
id: "Qwen/Qwen3-32B-fast",
name: "Qwen3 32B Fast",
reasoning: false,
input: ["text"],
cost: { input: 0, output: 0, cacheRead: 0, cacheWrite: 0 },
contextWindow: 128000,
maxTokens: 8192
}
]
}
}
}
}
```
Notes:
- Set `NEBIUS_API_KEY` in the environment or use `moltbot onboard --auth-choice nebius-api-key`.
- Model ref: `nebius/Qwen/Qwen3-32B-fast`.
- Available models include Qwen, DeepSeek, Llama, and GLM families.
### Synthetic (Anthropic-compatible)
Use Synthetic's Anthropic-compatible endpoint:

1
docs/providers/index.md
View File

@ -43,6 +43,7 @@ See [Venice AI](/providers/venice).
- [Amazon Bedrock](/bedrock)
- [Z.AI](/providers/zai)
- [Xiaomi](/providers/xiaomi)
- [Nebius Token Factory](/providers/nebius)
- [GLM models](/providers/glm)
- [MiniMax](/providers/minimax)
- [Venius (Venice AI, privacy-focused)](/providers/venice)

65
docs/providers/nebius.md Normal file
View File

@ -0,0 +1,65 @@
---
summary: "Use Nebius Token Factory with Moltbot"
read_when:
- You want Nebius models in Moltbot
- You need NEBIUS_API_KEY setup
---
# Nebius Token Factory
Nebius Token Factory provides OpenAI-compatible inference for frontier and open models. It offers a REST API compatible with OpenAI format and uses API keys for authentication. Create your API key in the [Nebius Token Factory console](https://tokenfactory.nebius.com/). Moltbot uses the `nebius` provider with a Nebius API key.
## Model overview
- **Qwen3 32B Fast**: 128000-token context window, fast inference.
- **DeepSeek V3 Fast**: Latest DeepSeek model with fast inference.
- **DeepSeek R1 Fast**: Reasoning model with chain-of-thought.
- **Llama 3.3 70B**: Strong general-purpose model with fast variant.
- **Qwen2.5 VL 72B**: Vision-language model.
- **Qwen2.5 Coder 7B Fast**: Code-specialized model.
- **GLM models**: GLM 4.7 and GLM 4.5 from Z.AI.
- Base URL: `https://api.tokenfactory.nebius.com/v1`
- Authorization: `Bearer $NEBIUS_API_KEY`
## CLI setup
```bash
moltbot onboard --auth-choice nebius-api-key
# or non-interactive
moltbot onboard --auth-choice nebius-api-key --nebius-api-key "$NEBIUS_API_KEY"
```
## Config snippet
```json5
{
env: { NEBIUS_API_KEY: "your-key" },
agents: { defaults: { model: { primary: "nebius/Qwen/Qwen3-32B-fast" } } },
models: {
mode: "merge",
providers: {
nebius: {
baseUrl: "https://api.tokenfactory.nebius.com/v1",
api: "openai-completions",
apiKey: "NEBIUS_API_KEY",
models: [
{
id: "Qwen/Qwen3-32B-fast",
name: "Qwen3 32B Fast",
reasoning: false,
input: ["text"],
cost: { input: 0, output: 0, cacheRead: 0, cacheWrite: 0 },
contextWindow: 128000,
maxTokens: 8192
}
]
}
}
}
}
```
## Notes
- Default model ref: `nebius/Qwen/Qwen3-32B-fast`.
- The provider is injected automatically when `NEBIUS_API_KEY` is set (or an auth profile exists).
- See [/concepts/model-providers](/concepts/model-providers) for provider rules.

118
src/agents/models-config.providers.ts
View File

@ -41,6 +41,17 @@ const XIAOMI_DEFAULT_COST = {
cacheWrite: 0,
};
const NEBIUS_BASE_URL = "https://api.tokenfactory.nebius.com/v1";
const NEBIUS_DEFAULT_MODEL_ID = "Qwen/Qwen3-32B-fast";
const NEBIUS_DEFAULT_CONTEXT_WINDOW = 128000;
const NEBIUS_DEFAULT_MAX_TOKENS = 8192;
const NEBIUS_DEFAULT_COST = {
input: 0,
output: 0,
cacheRead: 0,
cacheWrite: 0,
};
const MOONSHOT_BASE_URL = "https://api.moonshot.ai/v1";
const MOONSHOT_DEFAULT_MODEL_ID = "kimi-k2.5";
const MOONSHOT_DEFAULT_CONTEXT_WINDOW = 256000;
@ -370,6 +381,105 @@ export function buildXiaomiProvider(): ProviderConfig {
};
}
export function buildNebiusProvider(): ProviderConfig {
return {
baseUrl: NEBIUS_BASE_URL,
api: "openai-completions",
models: [
{
id: "Qwen/Qwen3-32B-fast",
name: "Qwen3 32B Fast",
reasoning: false,
input: ["text"],
cost: NEBIUS_DEFAULT_COST,
contextWindow: NEBIUS_DEFAULT_CONTEXT_WINDOW,
maxTokens: NEBIUS_DEFAULT_MAX_TOKENS,
},
{
id: "meta-llama/Meta-Llama-3.1-8B-Instruct-fast",
name: "Llama 3.1 8B Fast",
reasoning: false,
input: ["text"],
cost: NEBIUS_DEFAULT_COST,
contextWindow: NEBIUS_DEFAULT_CONTEXT_WINDOW,
maxTokens: NEBIUS_DEFAULT_MAX_TOKENS,
},
{
id: "meta-llama/Llama-3.3-70B-Instruct",
name: "Llama 3.3 70B",
reasoning: false,
input: ["text"],
cost: NEBIUS_DEFAULT_COST,
contextWindow: NEBIUS_DEFAULT_CONTEXT_WINDOW,
maxTokens: NEBIUS_DEFAULT_MAX_TOKENS,
},
{
id: "meta-llama/Llama-3.3-70B-Instruct-fast",
name: "Llama 3.3 70B Fast",
reasoning: false,
input: ["text"],
cost: NEBIUS_DEFAULT_COST,
contextWindow: NEBIUS_DEFAULT_CONTEXT_WINDOW,
maxTokens: NEBIUS_DEFAULT_MAX_TOKENS,
},
{
id: "deepseek-ai/DeepSeek-V3-0324-fast",
name: "DeepSeek V3 Fast",
reasoning: false,
input: ["text"],
cost: NEBIUS_DEFAULT_COST,
contextWindow: NEBIUS_DEFAULT_CONTEXT_WINDOW,
maxTokens: NEBIUS_DEFAULT_MAX_TOKENS,
},
{
id: "deepseek-ai/DeepSeek-R1-0528-fast",
name: "DeepSeek R1 Fast",
reasoning: true,
input: ["text"],
cost: NEBIUS_DEFAULT_COST,
contextWindow: NEBIUS_DEFAULT_CONTEXT_WINDOW,
maxTokens: NEBIUS_DEFAULT_MAX_TOKENS,
},
{
id: "Qwen/Qwen2.5-VL-72B-Instruct",
name: "Qwen2.5 VL 72B",
reasoning: false,
input: ["text", "image"],
cost: NEBIUS_DEFAULT_COST,
contextWindow: NEBIUS_DEFAULT_CONTEXT_WINDOW,
maxTokens: NEBIUS_DEFAULT_MAX_TOKENS,
},
{
id: "Qwen/Qwen2.5-Coder-7B-fast",
name: "Qwen2.5 Coder 7B Fast",
reasoning: false,
input: ["text"],
cost: NEBIUS_DEFAULT_COST,
contextWindow: NEBIUS_DEFAULT_CONTEXT_WINDOW,
maxTokens: NEBIUS_DEFAULT_MAX_TOKENS,
},
{
id: "zai-org/GLM-4.7-FP8",
name: "GLM 4.7 FP8",
reasoning: false,
input: ["text"],
cost: NEBIUS_DEFAULT_COST,
contextWindow: NEBIUS_DEFAULT_CONTEXT_WINDOW,
maxTokens: NEBIUS_DEFAULT_MAX_TOKENS,
},
{
id: "zai-org/GLM-4.5",
name: "GLM 4.5",
reasoning: false,
input: ["text"],
cost: NEBIUS_DEFAULT_COST,
contextWindow: NEBIUS_DEFAULT_CONTEXT_WINDOW,
maxTokens: NEBIUS_DEFAULT_MAX_TOKENS,
},
],
};
}
async function buildVeniceProvider(): Promise<ProviderConfig> {
const models = await discoverVeniceModels();
return {
@ -446,6 +556,14 @@ export async function resolveImplicitProviders(params: {
providers.xiaomi = { ...buildXiaomiProvider(), apiKey: xiaomiKey };
}
const nebiusKey =
resolveEnvApiKeyVarName("nebius") ??
resolveApiKeyFromProfiles({ provider: "nebius", store: authStore });
if (nebiusKey) {
providers.nebius = { ...buildNebiusProvider(), apiKey: nebiusKey };
}
// Ollama provider - only add if explicitly configured
const ollamaKey =
resolveEnvApiKeyVarName("ollama") ??

12
src/commands/auth-choice-options.ts
View File

@ -17,6 +17,7 @@ export type AuthChoiceGroupId =
| "moonshot"
| "zai"
| "xiaomi"
| "nebius"
| "opencode-zen"
| "minimax"
| "synthetic"
@ -114,6 +115,12 @@ const AUTH_CHOICE_GROUP_DEFS: {
hint: "API key",
choices: ["xiaomi-api-key"],
},
{
value: "nebius",
label: "Nebius",
hint: "Nebius Token Factory",
choices: ["nebius-api-key"],
},
{
value: "opencode-zen",
label: "OpenCode Zen",
@ -176,6 +183,11 @@ export function buildAuthChoiceOptions(params: {
label: "Xiaomi API key",
});
options.push({ value: "qwen-portal", label: "Qwen OAuth" });
options.push({
value: "nebius-api-key",
label: "Nebius API key",
hint: "Nebius Token Factory",
});
options.push({
value: "copilot-proxy",
label: "Copilot Proxy (local)",

69
src/commands/auth-choice.apply.api-providers.ts
View File

@ -17,6 +17,8 @@ import {
applyKimiCodeProviderConfig,
applyMoonshotConfig,
applyMoonshotProviderConfig,
applyNebiusConfig,
applyNebiusProviderConfig,
applyOpencodeZenConfig,
applyOpencodeZenProviderConfig,
applyOpenrouterConfig,
@ -32,6 +34,7 @@ import {
applyZaiConfig,
KIMI_CODE_MODEL_REF,
MOONSHOT_DEFAULT_MODEL_REF,
NEBIUS_DEFAULT_MODEL_REF,
OPENROUTER_DEFAULT_MODEL_REF,
SYNTHETIC_DEFAULT_MODEL_REF,
VENICE_DEFAULT_MODEL_REF,
@ -40,6 +43,7 @@ import {
setGeminiApiKey,
setKimiCodeApiKey,
setMoonshotApiKey,
setNebiusApiKey,
setOpencodeZenApiKey,
setOpenrouterApiKey,
setSyntheticApiKey,
@ -85,6 +89,8 @@ export async function applyAuthChoiceApiProviders(
authChoice = "zai-api-key";
} else if (params.opts.tokenProvider === "xiaomi") {
authChoice = "xiaomi-api-key";
} else if (params.opts.tokenProvider === "nebius") {
authChoice = "nebius-api-key";
} else if (params.opts.tokenProvider === "synthetic") {
authChoice = "synthetic-api-key";
} else if (params.opts.tokenProvider === "venice") {
@ -485,6 +491,69 @@ export async function applyAuthChoiceApiProviders(
return { config: nextConfig, agentModelOverride };
}
if (authChoice === "nebius-api-key") {
let hasCredential = false;
if (!hasCredential && params.opts?.token && params.opts?.tokenProvider === "nebius") {
await setNebiusApiKey(normalizeApiKeyInput(params.opts.token), params.agentDir);
hasCredential = true;
}
if (!hasCredential) {
await params.prompter.note(
[
"Nebius provides OpenAI-compatible inference for frontier and open models.",
"Get your API key at: https://tokenfactory.nebius.com/",
"Available models: zai-glm-7, zai-glm-5",
].join("\n"),
"Nebius",
);
}
const envKey = resolveEnvApiKey("nebius");
if (envKey) {
const useExisting = await params.prompter.confirm({
message: `Use existing NEBIUS_API_KEY (${envKey.source}, ${formatApiKeyPreview(envKey.apiKey)})?`,
initialValue: true,
});
if (useExisting) {
await setNebiusApiKey(envKey.apiKey, params.agentDir);
hasCredential = true;
}
}
if (!hasCredential) {
const key = await params.prompter.text({
message: "Enter Nebius API key",
validate: validateApiKeyInput,
});
await setNebiusApiKey(normalizeApiKeyInput(String(key)), params.agentDir);
}
nextConfig = applyAuthProfileConfig(nextConfig, {
profileId: "nebius:default",
provider: "nebius",
mode: "api_key",
});
{
const applied = await applyDefaultModelChoice({
config: nextConfig,
setDefaultModel: params.setDefaultModel,
defaultModel: NEBIUS_DEFAULT_MODEL_REF,
applyDefaultConfig: applyNebiusConfig,
applyProviderConfig: applyNebiusProviderConfig,
noteDefault: NEBIUS_DEFAULT_MODEL_REF,
noteAgentModel,
prompter: params.prompter,
});
nextConfig = applied.config;
agentModelOverride = applied.agentModelOverride ?? agentModelOverride;
}
return { config: nextConfig, agentModelOverride };
}
if (authChoice === "synthetic-api-key") {
if (params.opts?.token && params.opts?.tokenProvider === "synthetic") {
await setSyntheticApiKey(String(params.opts.token).trim(), params.agentDir);

79
src/commands/onboard-auth.config-core.ts
View File

@ -1,4 +1,9 @@
import { buildXiaomiProvider, XIAOMI_DEFAULT_MODEL_ID } from "../agents/models-config.providers.js";
import {
buildNebiusProvider,
buildXiaomiProvider,
NEBIUS_DEFAULT_MODEL_ID,
XIAOMI_DEFAULT_MODEL_ID,
} from "../agents/models-config.providers.js";
import {
buildSyntheticModelDefinition,
SYNTHETIC_BASE_URL,
@ -13,6 +18,7 @@ import {
} from "../agents/venice-models.js";
import type { MoltbotConfig } from "../config/config.js";
import {
NEBIUS_DEFAULT_MODEL_REF,
OPENROUTER_DEFAULT_MODEL_REF,
VERCEL_AI_GATEWAY_DEFAULT_MODEL_REF,
XIAOMI_DEFAULT_MODEL_REF,
@ -409,6 +415,77 @@ export function applyXiaomiConfig(cfg: MoltbotConfig): MoltbotConfig {
};
}
export function applyNebiusProviderConfig(cfg: MoltbotConfig): MoltbotConfig {
const models = { ...cfg.agents?.defaults?.models };
models[NEBIUS_DEFAULT_MODEL_REF] = {
...models[NEBIUS_DEFAULT_MODEL_REF],
alias: models[NEBIUS_DEFAULT_MODEL_REF]?.alias ?? "Qwen3 32B",
};
const providers = { ...cfg.models?.providers };
const existingProvider = providers.nebius;
const defaultProvider = buildNebiusProvider();
const existingModels = Array.isArray(existingProvider?.models) ? existingProvider.models : [];
const defaultModels = defaultProvider.models ?? [];
const hasDefaultModel = existingModels.some((model) => model.id === NEBIUS_DEFAULT_MODEL_ID);
const mergedModels =
existingModels.length > 0
? hasDefaultModel
? existingModels
: [...existingModels, ...defaultModels]
: defaultModels;
const { apiKey: existingApiKey, ...existingProviderRest } = (existingProvider ?? {}) as Record<
string,
unknown
> as { apiKey?: string };
const resolvedApiKey = typeof existingApiKey === "string" ? existingApiKey : undefined;
const normalizedApiKey = resolvedApiKey?.trim();
providers.nebius = {
...existingProviderRest,
baseUrl: defaultProvider.baseUrl,
api: defaultProvider.api,
...(normalizedApiKey ? { apiKey: normalizedApiKey } : {}),
models: mergedModels.length > 0 ? mergedModels : defaultProvider.models,
};
return {
...cfg,
agents: {
...cfg.agents,
defaults: {
...cfg.agents?.defaults,
models,
},
},
models: {
mode: cfg.models?.mode ?? "merge",
providers,
},
};
}
export function applyNebiusConfig(cfg: MoltbotConfig): MoltbotConfig {
const next = applyNebiusProviderConfig(cfg);
const existingModel = next.agents?.defaults?.model;
return {
...next,
agents: {
...next.agents,
defaults: {
...next.agents?.defaults,
model: {
...(existingModel && "fallbacks" in (existingModel as Record<string, unknown>)
? {
fallbacks: (existingModel as { fallbacks?: string[] }).fallbacks,
}
: undefined),
primary: NEBIUS_DEFAULT_MODEL_REF,
},
},
},
};
}
/**
* Apply Venice provider configuration without changing the default model.
* Registers Venice models and sets up the provider, but preserves existing model selection.

14
src/commands/onboard-auth.credentials.ts
View File

@ -116,6 +116,20 @@ export const ZAI_DEFAULT_MODEL_REF = "zai/glm-4.7";
export const XIAOMI_DEFAULT_MODEL_REF = "xiaomi/mimo-v2-flash";
export const OPENROUTER_DEFAULT_MODEL_REF = "openrouter/auto";
export const VERCEL_AI_GATEWAY_DEFAULT_MODEL_REF = "vercel-ai-gateway/anthropic/claude-opus-4.5";
export const NEBIUS_DEFAULT_MODEL_REF = "nebius/Qwen/Qwen3-32B-fast";
export async function setNebiusApiKey(key: string, agentDir?: string) {
// Write to resolved agent dir so gateway finds credentials on startup.
upsertAuthProfile({
profileId: "nebius:default",
credential: {
type: "api_key",
provider: "nebius",
key,
},
agentDir: resolveAuthAgentDir(agentDir),
});
}
export async function setZaiApiKey(key: string, agentDir?: string) {
// Write to resolved agent dir so gateway finds credentials on startup.

4
src/commands/onboard-auth.ts
View File

@ -9,6 +9,8 @@ export {
applyKimiCodeProviderConfig,
applyMoonshotConfig,
applyMoonshotProviderConfig,
applyNebiusConfig,
applyNebiusProviderConfig,
applyOpenrouterConfig,
applyOpenrouterProviderConfig,
applySyntheticConfig,
@ -35,12 +37,14 @@ export {
applyOpencodeZenProviderConfig,
} from "./onboard-auth.config-opencode.js";
export {
NEBIUS_DEFAULT_MODEL_REF,
OPENROUTER_DEFAULT_MODEL_REF,
setAnthropicApiKey,
setGeminiApiKey,
setKimiCodeApiKey,
setMinimaxApiKey,
setMoonshotApiKey,
setNebiusApiKey,
setOpencodeZenApiKey,
setOpenrouterApiKey,
setSyntheticApiKey,