Skip to main content

Baseten Provider Guide

Baseten is a Tier-2 catalog provider: OpenAI-wire-compatible with no behavioural quirks, so its entire integration is one JSON file (src/lib/providers/catalog/baseten.json) rather than hand-written code. That file is the source of truth for everything on this page.

Key Facts​

  • Provider id: baseten
  • Protocol: OpenAI-compatible (/chat/completions)
  • Base URL: https://inference.baseten.co/v1
  • Default model: zai-org/GLM-5.3-Flash
  • Models in catalog: 16
  • Streaming: supported
  • Tool calling: supported (native)
  • Structured output: supported
  • Embeddings: not supported
  • Billing: free-tier

Quick Start​

1. Get an API key​

  1. Visit: https://app.baseten.co/ and sign in or create a workspace
  2. Review the current Baseten billing and credit terms in the console before making requests
  3. Create a personal API key in the Baseten console
  4. Set BASETEN_API_KEY in your .env file

2. Configure​

export BASETEN_API_KEY=your-api-key
export BASETEN_MODEL=zai-org/GLM-5.3-Flash      # optional — overrides the default model
export BASETEN_BASE_URL=https://inference.baseten.co/v1   # optional — self-hosted or proxy

3. Use it​

import { NeuroLink } from "@juspay/neurolink";

const neurolink = new NeuroLink();

const result = await neurolink.generate({
  input: { text: "Explain context windows in one paragraph." },
  provider: "baseten",
  model: "zai-org/GLM-5.3-Flash",
});

console.log(result.content);
# CLI
npx @juspay/neurolink generate "Hello" --provider baseten

Per-request credentials work as they do for every provider:

await neurolink.generate({
  input: { text: "Hello" },
  provider: "baseten",
  credentials: { baseten: { apiKey: process.env.BASETEN_API_KEY } },
});

Models​

ModelContextVision$/M in · outNotes
openai/gpt-oss-120b125Kno$0.1 / $0.5OpenAI GPT-OSS 120B; general-purpose model with controllable reasoning
zai-org/GLM-4.7195Kno$0.6 / $2.2GLM 4.7; fast general-purpose model with 200K context and enhanced tool use
moonshotai/Kimi-K2.6256Kno$0.95 / $4Kimi K2.6; agentic and coding model for multi-step reasoning and tool use
deepseek-ai/DeepSeek-V4-Pro1Mno$1.74 / $3.48DeepSeek V4 Pro; 1M-context mixture-of-experts model for agentic workflows and coding
nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B198Kno$0.6 / $2.4NVIDIA Nemotron 3 Ultra; flagship reasoning and non-reasoning model for code and agentic execution
zai-org/GLM-5.21Mno$1.4 / $4.4GLM 5.2; 1M-context reasoning model
moonshotai/Kimi-K2.7-Code256Kno$0.95 / $4Kimi K2.7 Code; model for complex coding, code reasoning and long-horizon development
deepseek-ai/DeepSeek-V4-Flash-07311Mno$0.13 / $0.26DeepSeek V4 Flash 0731; fast, low-cost 1M-context mixture-of-experts model
thinkingmachines/inkling1Mno$1 / $4.05Thinking Machines Inkling; 1M-context reasoning model
zai-org/GLM-5.2-Fast1Mno$2.1 / $6.6GLM 5.2 Fast; 1M-context model
moonshotai/Kimi-K31Mno$3 / $15Kimi K3; 1M-context model
thinkingmachines/inkling-small1Mno$0.5 / $1.2Thinking Machines Inkling Small; 1M-context reasoning model
deepseek-ai/DeepSeek-V4-Pro-08131Mno$1.32 / $3.96DeepSeek V4 Pro 0813; dated 1M-context mixture-of-experts model for agentic workflows and coding
zai-org/GLM-5.3-Flash ⭐1Myes$0.15 / $0.5GLM 5.3 Flash; 1M-context reasoning model with live-verified image input
zai-org/GLM-5.31Mno$1.4 / $4.4GLM 5.3; 1M-context reasoning model
zai-org/GLM-5.3-Fast1Myes$2.1 / $6.6GLM 5.3 Fast; 1M-context reasoning model with image input

Fallback order when the default is unavailable: zai-org/GLM-5.3-Flash → zai-org/GLM-5.3 → zai-org/GLM-5.2.

Verification status​

Tier-2 onboarding requires evidence before a provider is accepted, and pnpm run verify:provider-onboarding gates it in CI. This is what the catalog records for Baseten:

ProbeResult
Rosterauthenticated GET /v1/models, HTTP 200, 2026-09-03
Auth rejectionHTTP 403, 2026-09-03
Live capability sweepnot run

⚠️ No live capability sweep is recorded for Baseten. The roster and auth behaviour were verified against the real API on the date above, but the capability flags come from the catalog declaration rather than from a measured end-to-end run. Treat them as the provider's stated behaviour.

Troubleshooting​

SymptomCauseFix
Invalid Baseten API keyBASETEN_API_KEY unset or wrongCheck the key at https://app.baseten.co/
Model not foundThe roster changed since 2026-09-03Pick a current id; catalog providers retire models without notice

See also​