Sophon ForgeNEW

Sophon Forge is the hosted inference service of the Sophon Platform. It gives you an OpenAI-compatible API to a managed set of language models, so you can run Sophon — or any app — without managing your own model keys or GPU infrastructure. Forge handles hosting, scaling, API keys, rate limits, and usage tracking for you.

What Forge offers

• OpenAI-compatible API — drop-in /v1/chat/completions, /v1/embeddings, and /v1/models endpoints. Point any OpenAI-compatible client at Forge by changing the base URL and API key.
• Managed models — a curated lineup of hosted models across speed/quality tiers, plus premium models on higher plans. Hosting is handled server-side, so models can be upgraded without changing your code.
• API key management — create, rotate, and revoke keys per workspace, with separate Live and Test environments. Keys are shown once and stored hashed.
• Workspaces — keys, usage, and plans are scoped to a workspace, so teams can share access with Owner/Member roles.
• Usage tracking — a dashboard showing current-month token usage by model and key, plus historical daily aggregates.
• Plans & quotas — monthly token allowances and per-minute request limits enforced per plan, with rate-limit headers on every response.

Plans & limits

Every account starts on the Free plan automatically. Higher plans raise the monthly token allowance and request rate, and unlock premium models.

Current pricing and plan availability are shown on the Forge dashboard.

Getting started

1. Create an account

Sign up at platform.sophon.buildersoft.io and verify your email. A personal workspace is created for you automatically.

2. Create an API key

1. Open your workspace and go to Forge → API Keys
2. Click Create API Key, give it a name, and choose Live or Test
3. Copy the key (format forge_sk_live_...) — it is shown only once

3. Call the API

Forge speaks the OpenAI protocol. Use your key as a Bearer token against the Platform base URL:

curl https://api.platform.sophon.buildersoft.io/v1/chat/completions \
  -H "Authorization: Bearer forge_sk_live_..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "<model-id>",
    "messages": [{ "role": "user", "content": "Hello" }]
  }'

List the model ids available on your plan with GET /v1/models, or browse them in the dashboard. Responses include rate-limit headers — X-RateLimit-Remaining, X-RateLimit-Reset, and X-Forge-Tokens-Remaining — so clients can back off before hitting a limit.

Use Forge in Sophon

Forge is available in the Sophon app as the sophon-forge model provider, so you can run your agents on Forge-hosted models instead of bringing your own provider key.

1. In the Dashboard, go to Settings → Models & Providers → Add Provider
2. Choose Sophon Forge
3. Paste your Forge API key (and, if prompted, the base URL https://platform.sophon.buildersoft.io)
4. Test the connection, then set its priority and budget

See the full provider list in Supported Providers and general model setup in Configuration.