Skip to main content
Watch: I Built an AI App Without Leaving Databricks

API compatibility - 9 documented limits

Fireworks AI API compatibility

Fireworks AI documents an OpenAI-compatible API at https://api.fireworks.ai/inference/v1 (chat/completions and completions). Its Anthropic-compatible base URL, the value for ANTHROPIC_BASE_URL, is https://api.fireworks.ai/inference. The docs list 9 limitations: 1 for the OpenAI-compatible layer and 8 for the Anthropic-compatible layer.

Verified against Fireworks AI's own documentation. Not documented means the docs read that day do not say; it never means No.

Compare every provider in the API compatibility matrix.

OpenAI-compatible API

Supported
Base URL
https://api.fireworks.ai/inference/v1
Endpoints
  • chat/completions
  • completions
Auth
Authorization: Bearer $API_KEY (Fireworks API key)
Notes
Model must be a Fireworks model resource name (accounts/fireworks/models/...). Usage stats are returned on streaming responses too, in the final chunk.

From the docs

You can use the OpenAI Python client library to interact with Fireworks. This makes migration of existing applications already using OpenAI particularly easy.
docs.fireworks.ai/tools-sdks/openai-compatibility

Anthropic-compatible API

Supported
ANTHROPIC_BASE_URL
https://api.fireworks.ai/inference
Auth
ANTHROPIC_AUTH_TOKEN (sent as Bearer); docs say not to use ANTHROPIC_API_KEY with Claude Code
Regional and alternate URLs
  • Defaulthttps://api.fireworks.ai/inference
  • US Only Serverlesshttps://us.api.fireworks.ai/inference
Notes
Base URL has no /v1 suffix; the SDK appends /v1/messages. Supported for serverless and on-demand deployments, not direct route endpoints. Only POST /v1/messages is served.

From the docs

The base URL for the Anthropic SDK is https://api.fireworks.ai/inference (without the /v1 suffix). The SDK appends /v1/messages automatically.
docs.fireworks.ai/tools-sdks/anthropic-compatibility

Documented limitations

Parameters, content types and features Fireworks AI's docs say its compatibility layers ignore, reject or change, each with the line it comes from. A layer with nothing listed may still differ in ways the docs do not spell out.

OpenAI-compatible layer1

  • max_tokens auto-truncated when context exceeded (OpenAI errors instead)

    If the length of prompt or messages plus max_tokens is higher than the model's context window, max_tokens will be adjusted lower accordingly. OpenAI returns an invalid request error in this situation.
    docs.fireworks.ai/tools-sdks/openai-compatibility

Anthropic-compatible layer8

Example from Fireworks AI's docs

Copied verbatim from the provider's documentation, placeholders included.

Provider example
from openai import OpenAI

client = OpenAI(
    base_url="https://api.fireworks.ai/inference/v1",
    api_key="<YOUR_FIREWORKS_API_KEY>",
)

chat_completion = client.chat.completions.create(
    model="accounts/fireworks/models/llama-v3p1-8b-instruct",
    messages=[
        {
            "role": "system",
            "content": "You are a helpful assistant.",
        },
        {
            "role": "user",
            "content": "Say this is a test",
        },
    ],
)

Newsletter

Know when a provider changes its API

One email per week. New models, new endpoints and the compatibility changes that break a base URL, and what they mean for your stack.