LoginSign Up
← Back to all sparks
B

Baseten

AI-ASSISTANTS
Velocity7.5

AI model deployment and inference platform for running ML models in production.

www.baseten.co ↗

Baseten is selling to the labs that build models, not just the developers who call them.

model-apisinference-servingthroughput-tieringmodel-labsworkspace-governancecatalog-churn
◆Current state
The catalog turns over constantly — DeepSeek V4 Pro 0813, Inkling and Inkling Small, Kimi K3, GLM 5.2 Fast in, and GLM 5.1, GLM 5, Kimi K2.5 and Nemotron Super 120B deprecated — all reachable through the same OpenAI-compatible endpoint with dedicated deployments for larger workloads. Two releases break that pattern. Baseten for Model Labs packages the serving stack as infrastructure a lab can adopt instead of building its own, and the Fast tier debuts with GLM 5.2 Fast: identical weights on dedicated capacity tuned for sustained per-user throughput. Workspace governance fills in alongside — org-scoped key administration, programmatic logs and metrics, and a GPU usage view for admins.
◆Where it's heading
Baseten is working both sides of the market at once. Toward developers, model choice is being commoditised into interchangeable catalog entries while serving characteristics become the thing that is actually priced. Toward labs, the pitch is that distribution and serving are someone else's problem. Those converge on the same position: whoever owns the endpoint owns the relationship, regardless of who trained the weights. The governance releases are the unglamorous prerequisite for the larger accounts that position requires.
◆Prediction
Expect the Fast tier to expand beyond GLM 5.2 to the models agentic workloads lean on hardest, and the deprecation cadence to continue thinning older catalog entries as newer ones land.

◆Recent moves

  1. 5d ago

    DeepSeek V4 Pro 0813 available on Baseten

    DeepSeek V4 Pro 0813 becomes callable through Model APIs on the standard OpenAI-compatible endpoint. Routine catalog maintenance — and the speed of these additions is precisely what makes the model layer interchangeable and pushes Baseten's differentiation onto serving tiers and the labs business.

    View source ↗
  2. 19d ago

    Inkling Small available on Baseten

    Inkling Small joins the Model APIs catalog through the standard OpenAI-compatible endpoint, with dedicated deployments available for larger workloads. A routine catalog addition, arriving two weeks after Inkling itself.

    View source ↗
  3. 20d ago

    Introducing Baseten for Model Labs

    ⚡ SPARK

    Baseten starts selling to the labs that build models, not just the developers who call them. This is the clearest directional move in the current batch and reframes what the rest of the platform work is for.

    View source ↗
  4. 22d ago

    Kimi K3 available on Baseten

    Kimi K3 becomes callable through the Model APIs on the same endpoint-plus-dedicated-deployment pattern as every other catalog addition. Steady catalog maintenance rather than a change in what the platform does.

    View source ↗
  5. 26d ago

    GLM 5.2 Fast available on Baseten

    ⚡ SPARK

    The Fast tier debuts here, establishing that serving capacity — not just model choice — is now something customers select and pay for. It sets up the tiering that later catalog additions can slot into.

    View source ↗
  6. 26d ago

    API key management keys

    A new org-scoped key type lets teams automate key administration — creating team keys at any permission level, and listing or revoking team and personal keys programmatically. Part of the governance stream aimed at larger workspaces.

    View source ↗