> Baseten
Baseten offers Model APIs, hosted open-weight models behind an OpenAI Chat Completions endpoint at inference.baseten.co with a beta Anthropic Messages endpoint, alongside dedicated GPU deployments and training billed per minute. Model API rates range from $0.10/1M input and $0.50/1M output for GPT OSS 120B to $3.00/1M input and $15.00/1M output for Kimi K3, with cached input billed at roughly a tenth of the input rate. The Basic plan is $0 per month pay-as-you-go, and the pricing FAQ states new accounts come with credits but publishes no amount.
// pros
- Basic plan carries no platform fee ($0 per month, pay as you go) and includes dedicated deployments, Model APIs, training, fast cold starts, SOC 2 Type II and HIPAA compliance.
- Model API rate limits are published by tier: Basic unverified 15 RPM / 100,000 TPM, Basic verified 120 RPM / 500,000 TPM, Pro 120 RPM / 1,000,000 TPM.
- Prompt caching is automatic and discounted, for example DeepSeek V4 Flash 0731 at $0.028/1M cached against $0.13/1M uncached, with session affinity via the x-session-affinity header to keep prefixes warm.
- Dedicated GPU billing is per minute (H100 $0.10833/min, B200 $0.16633/min, A100 80GB $0.06667/min, L4 $0.01414/min), finer granularity than hourly instance billing.
- The Model API catalogue includes 1048k-context models (GLM 5.3, DeepSeek V4 Pro 0813, Kimi K3) and Fast variants with separate slugs, pricing and rate limits for higher sustained throughput.
- Workspace monthly budgets can be enforced so Model API requests are rejected at the cap, and usage and cost are available through REST, CLI and Prometheus.
// cons
- Unverified Basic accounts are limited to 15 RPM and 100,000 TPM, below what most agent or batch workloads need; the step up to 120 RPM requires account verification.
- No sign-up credit amount is published: the FAQ only says new accounts come with credits, so free runway cannot be planned or compared.
- Model API capacity is best-effort: the docs note a 529 Overloaded response can occur even while you remain inside your rate limits, so retry and backoff logic is mandatory.
- Anthropic Messages API support is beta, with behaviour that may change before general availability, and production use is directed to Chat Completions.
- Usage history is bounded: Model API usage data is retained 92 days and cost data begins on 5 August 2026 at 20:45 UTC, which limits long-range cost auditing.
Use Baseten when you want a hosted open-weight Model API with an OpenAI-compatible surface on a pay-as-you-go $0 plan, and pair it with a free-tier provider for bulk prototyping while the account is still unverified and limited to 15 RPM.
written by Mike Fleming, Uprouter editorial · verified against official sources September 16, 2026
Free tier present — value not quantified
This provider offers a free tier, but its terms don’t convert cleanly into a USD figure (e.g. request-count limits or capacity-dependent pools). We refuse to print a misleading zero here. See the note below and the official pricing page for the exact limits.
note: The official pricing FAQ states that new Baseten accounts come with credits to experiment with deployments, but no dollar amount, expiry or conditions are published (checked 2026-09-16).
| Plan | Type | Monthly | Input /1M | Output /1M | Note |
|---|---|---|---|---|---|
| Free tier (source-reported) | free | — | — | — | New workspaces get ~$30 trial credits, billed by compute time; Startup plan has no platform fee. Converts to pay-as-you-go once credits are spent. |
| Free Trial Credits | free | — | — | — | $30 one-time credit for new accounts to deploy and test models. |
| Compute (PAYG) | payg | — | — | — | Billed by GPU compute time; varies by instance type and duration. |
Pricing normalized from public sources — always verify with the provider.
// model prices
| Model | Input /1M | Output /1M | Context |
|---|---|---|---|
| Any supported model in their library (billed by compute; no fixed free-model list) | — | — | — |
$ Does Baseten have a free tier?
Yes — Baseten offers a free tier. Always confirm current limits on Baseten's official pricing page — free tiers change without notice.
$ How much does Baseten cost per 1M tokens?
Baseten does not publish simple per-1M-token pricing (it may bill per request, per output, or via a subscription plan). Check their official pricing page for exact figures.
$ Is Baseten safe to use?
Uprouter rates Baseten as low risk. Medium. The Basic plan is free to hold but fully metered, with published tier limits that cap prototyping throughput at 15 RPM and 100,000 TPM until the account is verified. The only free grant is an unspecified credits line in the pricing FAQ, so nothing can be planned around it. Capacity is explicitly best-effort, with 529 Overloaded documented inside rate limits, and budget enforcement covers Model API requests only - not Dedicated Inference or Training - so those spend paths need separate monitoring. Risk ratings are editorial, evidence-based, and never influenced by affiliate relationships.
Raw community sentiment, shown unblended. Votes never feed into objective data fields like price, risk, or status — and never into sort order.
No published reviews yet — be the first.
// more_informationprice history · live status · risk & confidence · code snippets
price_history
No recorded changes yet.
live_status
current: upNo probes recorded yet.
Uprouter runs its own probes; best-effort, not a guarantee. Probe results depend on our network location and cadence — for production, always use the provider’s own status page.
risk_&_confidence
Medium. The Basic plan is free to hold but fully metered, with published tier limits that cap prototyping throughput at 15 RPM and 100,000 TPM until the account is verified. The only free grant is an unspecified credits line in the pricing FAQ, so nothing can be planned around it. Capacity is explicitly best-effort, with 529 Overloaded documented inside rate limits, and budget enforcement covers Model API requests only - not Dedicated Inference or Training - so those spend paths need separate monitoring.
Confidence reflects how much verifiable evidence (official docs, probe history, community reports) backs this entry. Risk is our editorial assessment of operator reliability and terms — not financial advice.
code_snippets
# Baseten's own API — get a key from their dashboard
curl https://YOUR-PROVIDER-ENDPOINT/chat/completions \
-H "Authorization: Bearer THEIR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "any-supported-model-in-their-library-billed-by-compute-no-fixed-free-model-list",
"messages": [{ "role": "user", "content": "Hello" }]
}'Direct to the provider — Uprouter receives nothing; verify their current pricing before purchasing.
Ranked by neutral similarity signals only — live status, free-tier shape and free-credit proximity. Commercial relationships never influence this list.