CAMELSTREAM · AGENT GUIDE
Give Codex unlimited frontier intelligence.
Codex supports named providers over its native Responses API. Add camelStream as one and your coding sessions run on frontier models for $5 a month per stream, with the meter gone.
Setup
One provider, one profile.
Export your Stream key
Set CAMEL_API_KEY in the shell that launches Codex. Keep the raw key out of your TOML files.
Add the provider and profile
Register camel_stream in ~/.codex/config.toml, then create ~/.codex/camel-stream.config.toml with the Stream model selection.
Launch the Stream profile
Run codex --profile camel-stream in a test repository. Codex will use Stream's /v1/responses endpoint.
# Add to ~/.codex/config.toml
[model_providers.camel_stream]
name = "camelStream"
base_url = "https://stream.camelai.com/v1"
env_key = "CAMEL_API_KEY"
wire_api = "responses"
# Create ~/.codex/camel-stream.config.toml
model = "auto"
model_provider = "camel_stream"stream.camelai.com/v1
auto
Responses
env CAMEL_API_KEY
Why Codex on Stream
Native transport, separate profile.
Codex speaks the Responses API, and so does Stream, so there's no compatibility shim in the middle. The provider lives in your Codex config as its own profile: run codex --profile camel-stream when you want the flat rate, and your regular OpenAI setup stays exactly where it was.
Responses API, end to end
Stream serves Codex over /v1/responses, the transport Codex prefers, with streaming and tool calls intact. No compatibility plugin in the middle.
Profiles keep it clean
The camel-stream profile is additive. Nothing about your default provider, keys, or unrelated Codex settings is overwritten.
Loops without a ledger
Repository exploration, test runs, retries: coding-agent traffic is bursty by nature. On Stream the bursts queue instead of billing.
The deal
What every stream includes.
Frontier models only
You never pick a model; keeping the fleet smart is our job. Every model we serve scores 70% or higher on Terminal-Bench 2.1, or 50 or higher on the Artificial Analysis Intelligence Index, and always at its newest public release.
260K context, guaranteed
Every request gets at least a 260K-token window, and more when the model serving you supports it. Beyond the guaranteed window we compact the middle of the conversation first; your original task and latest turns stay intact.
One generation per stream
Unlimited works because each stream runs one generation at a time. Extra requests wait in a visible queue instead of raising your bill; add streams for parallel capacity.
$5 a month, flat
No token metering, no overage charges, cancel anytime. Sourcing efficiently across the fleet is what keeps the price flat.
The fleet changes as new models launch. The floor doesn't. The full spec sheet lives on the Stream page, and speed numbers are published targets, not a service-level agreement — the fine print is in the terms.
Troubleshooting
Check these first.
Most connection issues come down to the endpoint, API mode, model ID, or key. Start here before anything else.
The Codex base URL includes /v1: https://stream.camelai.com/v1.
Set wire_api to responses, not chat or chat_completions.
Put the provider in your user Codex config; project-level provider settings are ignored.
If authentication fails, confirm CAMEL_API_KEY is available in the same shell that starts Codex.
Frequently asked questions
Codex, answered.
More guides
Point another agent at Stream.
Give Codex the whole stream.
Unlimited frontier intelligence over the API format Codex already speaks, for $5 a month per stream.
Get your API key