Metered access · Sources reviewed 2026-07-25

Together AI free tier

What Together AI publishes about its free allowance, which models it covers, and how to point an existing tool at it.

What Together AI's free tier allows

Together applies dynamic per-model rate limits that rise with sustained successful traffic and fall when traffic drops, and states plainly that there are no fixed per-model limits published. Requests above your dynamic rate return 429 with x-ratelimit-reset; requests at or below it that still fail return 503. No standing free allowance is documented on the rate-limits page.

Published limits
Dynamic, no fixed numbers
Free access type
Metered access
Credit card
Not required
Protocol
OpenAI-compatible at https://api.together.xyz/v1
Lifecycle
Active
Sources reviewed
2026-07-25

Because limits are usage-derived, a new account cannot know its ceiling in advance; read the 429 response headers instead.

Where Together AI sits in this catalog

Together AI is one of 6 metered access entries among the 25 providers detailed here. It asks for no credit card, which is true of 23 of the 25. It publishes no fixed request count, as 16 of the 25 do not.

Compared withPublished limitsCardEndpoint
Together AI (this page) Dynamic, no fixed numbers Not required api.together.xyz
Nebius Token Factory 60 requests per minute Not required api.tokenfactory.nebius.com
Perplexity API 50 requests per minute Not required api.perplexity.ai
DeepInfra Enforced but not published Not required api.deepinfra.com

Models Together AI lists as free

  • meta-llama/Llama-3.3-70B-Instruct-Turbo — Llama, also free at 8 other providers here
  • openai/gpt-oss-120b — gpt-oss, also free at 9 other providers here
  • Qwen/Qwen3-Coder-480B-A35B-Instruct-Turbo — Qwen, also free at 8 other providers here

9 other providers here also host the gpt-oss family on free terms, and the terms are not the same: GroqCloud, SambaNova Cloud, Cloudflare Workers AI, Hugging Face Inference Providers, Fireworks AI, Ollama Cloud, Cerebras Inference, Vercel AI Gateway, Scaleway Generative APIs. 8 other providers here also host the Llama family on free terms, and the terms are not the same: GroqCloud, SambaNova Cloud, Cloudflare Workers AI, Fireworks AI, IBM watsonx.ai, Nebius Token Factory, DeepInfra, Scaleway Generative APIs. 8 other providers here also host the Qwen family on free terms, and the terms are not the same: Cloudflare Workers AI, SiliconFlow, Alibaba Cloud Model Studio, Ollama Cloud, Nebius Token Factory, DeepInfra, Chutes, Scaleway Generative APIs.

Checking a Together AI key

The CORS preflight allows any origin to send an Authorization header, so a browser can call this endpoint directly. That was measured against https://api.together.xyz on 2026-07-25.

So a key can be checked in the browser checker without installing anything. If you would rather not paste a key into a web page, this does the same thing:

#!/usr/bin/env sh
set -eu

: "${TOGETHER_API_KEY:?Set TOGETHER_API_KEY to a key you created yourself at the provider console}"

# Prints the models this key can reach, then the HTTP status on its own line so
# 200, 401, and 429 stay distinguishable.
curl --silent --show-error \
  --write-out '\nHTTP %{http_code}\n' \
  --header "Authorization: Bearer $TOGETHER_API_KEY" \
  'https://api.together.xyz/v1/models'

Using a Together AI key

Together AI serves the OpenAI chat completions protocol at https://api.together.xyz/v1. Point any client that accepts a custom base URL at it, and let the client read the key from TOGETHER_API_KEY so the value never lands in a config file:

# Merge this snippet into ~/.codex/config.toml.
# Project-level config cannot select a custom model provider.
model = "meta-llama/Llama-3.3-70B-Instruct-Turbo"
model_provider = "together"

[model_providers.together]
name = "Together AI OpenAI-compatible gateway"
base_url = "https://api.together.xyz/v1"
env_key = "TOGETHER_API_KEY"
wire_api = "responses"

Questions about the Together AI free tier

Does Together AI ask for a credit card?

No. Together AI is one of 23 providers here that hand out free access without a card, so the only cost of trying it is the signup.

What are Together AI's free-tier rate limits?

Together AI publishes no single free-tier number: its limits are dynamic, no fixed numbers. This catalog leaves that blank rather than guessing a figure.

What happens when a Together AI key hits the limit?

The endpoint answers 429. That is a statement about the request that was refused, not about how much quota is left; only a reset header from the provider tells you when it clears.

Is the Together AI free tier going away?

Nothing in Together AI's own documentation says so as of 2026-07-25. Because limits are usage-derived, a new account cannot know its ceiling in advance; read the 429 response headers instead.

Can Together AI be used with Codex, Cline, or Continue?

Yes. Together AI serves the OpenAI chat completions protocol at https://api.together.xyz/v1, so any client that accepts a custom base URL can use it. Claude Code is the exception, because it speaks the Anthropic protocol instead.

Official sources

Every figure above was read from those pages on 2026-07-25. Where a provider states a limit only inside a console, this catalog records that fact instead of a number. See the methodology for how an entry gets in and how it gets corrected.