Backboard Now Connects You to 2,200 LLMs

What's New

Backboard now gives developers access to 2,200+ LLMs through a single API — enabling flexible, cost-efficient multi-model routing across every major provider without managing individual integrations.

Supported Providers

• OpenAI — GPT-4, GPT-4 Turbo, GPT-3.5 Turbo, and all embedding models.

• Anthropic — Claude 3.5 Sonnet, Claude 3 Opus, Haiku.

• Google — Gemini Pro, Gemini Ultra, PaLM 2.

• Meta — Llama 2 (7B, 13B, 70B), Code Llama.

• Mistral — Mistral 7B, Mixtral 8x7B, Mistral Large.

• Cohere — Command, Command Light, Embed models.

• And 50+ more providers via unified API access.

Key Features

• Single endpoint: One API call, any model — no provider-specific SDK required.

• Automatic fallback: If a provider is down, requests automatically route to alternatives.

• Unified billing: One invoice for all model usage across all providers.

• Cost tracking: Per-request cost breakdowns regardless of which provider served it.

How to Use It

Pass any supported model name to the model parameter in your chat completion request. Backboard handles authentication, routing, and response normalization automatically. No code changes needed when switching providers.