togetherlinkRun Prime Agent on Together AI Models
Keep Prime Agent's persistent REPL, subagents, and long-running sessions while its selected model is served by Together AI.
By TogetherLink
The short answer
TogetherLink registers its curated catalog as an OpenAI-compatible Prime provider for this launch. Prime Agent still owns tools, sessions, the Python control environment, subagents, and background execution.
Start a first Together-backed session
Install Prime Agent
Use the upstream project's supported installer. See the Prime Agent quickstart for platform-specific options.
curl -fsSL https://app.primeintellect.ai/prime-agent/install.sh | shInstall and configure TogetherLink
Save a Together API key once. The selected model is billed by Together AI, not by the harness vendor.
curl -fsSL https://togetherlink.vercel.app/install.sh | bashtogetherlink configureLaunch Prime Agent
tprimetogetherlink --model zai-org/GLM-5.2 primeWhat TogetherLink changes—and what it leaves alone
| Area | Behavior |
|---|---|
| Model requests | The harness calls Together AI directly through its OpenAI-compatible API. |
| Together key | Provided only to the launched process; it is not written into the generated provider metadata. |
| Harness state | Prime's home, sessions, skills, memories, extensions, workers, and background services remain native and available. |
| Provider setup | A credential-free provider extension is stored under ~/.togetherlink/prime-agent and activated only for the TogetherLink launch. |
| Usage reporting | TogetherLink records launch lifecycle only. Token and cost totals are not available locally for this direct harness. |
Prime Agent is powerful, but it is not a security sandbox
Prime Agent can execute model-generated Python and project commands with your user permissions. Use a reviewable worktree or external sandbox for untrusted repositories and instructions.
Together changes the model, not the workflow
Persistent RLM
Use Prime's IPython control environment and recursive subagents with the selected Together model.
Long-running work
Detach, reattach, schedule, and continue daemon-backed sessions through Prime's native workflow.
Prime extensions
User extensions and ordinary Prime arguments remain available after TogetherLink pins the provider and model.
Model choice
Switch among the five curated Together models without editing Prime's normal configuration.
5 curated models, with Kimi K3 as the default
To override the model, put --model before the harness name. Arguments after the harness name belong to the underlying tool.
moonshotai/Kimi-K3Default · vision · reasoning · 1M context
zai-org/GLM-5.2text · reasoning · 512K context
deepseek-ai/DeepSeek-V4-Pro-0813text · reasoning · 1M context
Qwen/Qwen3.7-Maxvision · reasoning · 1M context
deepseek-ai/DeepSeek-V4-Flash-0731text · reasoning · 1M context
Verify the launch before starting work
TogetherLink prints a launch line before the harness opens. For the default command, look for:
togetherlink ▸ Launching Prime Agent with Together AI.
Launching the upstream binary directly skips TogetherLink and uses that tool's normal provider configuration.
FAQ
Can Prime Agent use Together AI models?+
Yes. TogetherLink activates a credential-free provider extension for the launch, pins Prime Agent to Together AI, and supplies the key only at runtime.
Does this replace Prime Agent's provider settings?+
No. The Together provider is activated with a launch flag and the generated extension lives under TogetherLink's own home. Prime Agent's normal home and provider configuration are not rewritten.
Do background agents keep working?+
Prime Agent's daemon-backed sessions and workers remain Prime features. TogetherLink stores stable provider metadata under its own home so a detached worker can resolve the provider without storing your API key there.
Can I use autonomous mode?+
Yes. Put Prime Agent flags after the harness name, for example: tprime --autonomous. TogetherLink strips conflicting provider, model, model-list, and API-key overrides but preserves other Prime Agent arguments.
Does TogetherLink report Prime token cost?+
No. Prime Agent talks to Together directly, so TogetherLink records session lifecycle but does not see the inference stream or produce local token and cost totals.