Provider Dossier · GPU Cloud
Baseten
Production dedicated deployments with strong autoscaling.
The Datum Index verdict
Recommended with confidence
Ranked No. 9 of 20 in our composite reputation index, and currently improving.
≈ 4.2 / 5 aggregate · 260 reviewer notes
Composite index · reviewed 10 July 2026
Findings
- Baseten holds a Datum Index composite reputation of 84 / 100 (≈ 4.2 / 5), ranked No. 9 of 20.
- It is a gpu cloud based in San Francisco, USA, founded 2019, offering open-weight model access.
- Its flagship offering is Dedicated + Model APIs, exposed through an OpenAI-compatible API.
- Indicative pricing is $0.70 / 1M tokens in / $0.90 / 1M tokens out (medium confidence).
- Reference throughput is ~116 tok/s with ~2040 ms time-to-first-token; stated uptime is 99.6%.
Baseten enters our index at No. 9, a placement that reads as recommended with confidence. The composite rests on 260 reviewer notes gathered across the public record, and it is best understood not as a score out of ten but as a standing among peers. The trend line is pointed upward: reviewer sentiment has strengthened over recent quarters.
In its favour
Production dedicated deployments with strong autoscaling.
On the numbers that flatter it: open-weight access, 3 named compliance attestations, and a reference throughput of ~116 tokens per second.
Points of concern
Dedicated capacity costs more than shared per-token tiers.
The caveat a buyer should price in before committing production traffic. As with every entry, we weigh it against the field rather than against perfection.
Incident & Uptime Note
The record of being there
Baseten states an availability of 99.6%. Held over a full year, that headline implies on the order of 35.0 hours of cumulative unavailability — a useful sense of scale, though real incidents cluster rather than spread evenly, and a single bad afternoon can outweigh a quiet quarter.
Observatory Panel
Latency & throughput, in distribution
Monthly panels across the models Baseten serves seed
GLM 5.2
| Window | TTFT ms — mean · p50 · p90 | Throughput tok/s — mean · p50 · p90 | ||||
|---|---|---|---|---|---|---|
| May 2026 | ~2340 | ~2130 | ~3920 | ~108 | ~111 | ~127 |
| June 2026 | ~2320 | ~2160 | ~4110 | ~101 | ~109 | ~128 |
| July 2026 | ~2170 | ~2030 | ~3210 | ~113 | ~117 | ~126 |
Llama 3.3 70B (dedicated)
| Window | TTFT ms — mean · p50 · p90 | Throughput tok/s — mean · p50 · p90 | ||||
|---|---|---|---|---|---|---|
| May 2026 | ~1150 | ~1010 | ~1710 | ~86 | ~89 | ~97 |
| June 2026 | ~1060 | ~960 | ~1550 | ~91 | ~94 | ~100 |
| July 2026 | ~1240 | ~1060 | ~2290 | ~83 | ~85 | ~99 |
All figures are single-stream, per-request rates as one user would see them — never aggregate accelerator throughput. Serving modes are not comparable with one another: shared serverless APIs batch many tenants per accelerator, dedicated and self-served deployments hand one tenant the whole card, and specialist silicon is a regime of its own. Read each figure within its mode.
No field returns on record for Baseten yet — the ledger above opens as soon as the first practitioner submission clears verification.
Key Facts
The record, in brief
| Category | GPU Cloud |
|---|---|
| Headquarters | San Francisco, USA |
| Founded | 2019 |
| Model access | Open-weight |
| Flagship / reference | Dedicated + Model APIs · Llama 3.3 70B |
| OpenAI-compatible API | Yes |
| Indicative price (in / out) | $0.70 / 1M tokens / $0.90 / 1M tokens medium |
| Reference throughput | ~116 tok/s |
| Reference TTFT | ~2040 ms |
| Stated uptime | 99.6% |
| Compliance | SOC 2 Type II · HIPAA · GDPR |
| Composite reputation | 84 / 100 · ≈ 4.2 / 5 · 260 notes |
Reviewer Notes
What the field says
The 260 notes behind Baseten's standing are drawn from the accumulated public verdict of practitioners — the recurring praises and the recurring gripes — rather than from any single survey. Two themes dominate. Admirers return to one point above all: Production dedicated deployments with strong autoscaling. Detractors return to another: Dedicated capacity costs more than shared per-token tiers.
We publish neither individual reviews nor reviewer identities; the composite is an editorial synthesis, and the reviewer-note count is an order-of-magnitude indication of how much public signal informs it.
Related dossiers · see also
Peer entries in the same or adjacent category, for comparison against Baseten's standing.