The reputation index for AI infrastructure Vol. I · No. 1 · 10 July 2026

DatumIndex

Sine ira et studio — the record, without favour

Provider Dossier · Aggregator

DeepInfra

Aggressive per-token pricing on popular open models.

The Datum Index verdict

Recommended with confidence

Ranked No. 14 of 20 in our composite reputation index, and currently holding steady.

81 / 100

≈ 4.1 / 5 aggregate · 305 reviewer notes

Composite index · reviewed 10 July 2026

Findings

  1. DeepInfra holds a Datum Index composite reputation of 81 / 100 (≈ 4.1 / 5), ranked No. 14 of 20.
  2. It is a aggregator based in Palo Alto, USA, founded 2022, offering open-weight model access.
  3. Its flagship offering is Pay-per-token open models, exposed through an OpenAI-compatible API.
  4. Indicative pricing is $0.35 / 1M tokens in / $0.40 / 1M tokens out (medium confidence).
  5. Reference throughput is ~13 tok/s with ~2250 ms time-to-first-token; stated uptime is 99.5%.
Figure · Dossier No. 14 Standing relative to the field
  • DeepInfra 81
  • Field mean (20 providers) 83
  • Index leader (OpenAI) 91
Composite index (0–100), ranked No. 14 of 20. Source: Datum Index, reviewed 10 July 2026. Editorial composite — indicative, not a measured benchmark.

DeepInfra enters our index at No. 14, a placement that reads as recommended with confidence. The composite rests on 305 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 flat: reviewer sentiment has held steady over recent quarters.

In its favour

Aggressive per-token pricing on popular open models.

On the numbers that flatter it: open-weight access, 2 named compliance attestations, and a reference throughput of ~13 tokens per second.

Points of concern

Fewer enterprise features than larger platforms.

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

DeepInfra states an availability of 99.5%. Held over a full year, that headline implies on the order of 43.8 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 DeepInfra serves medium

Llama 3.3 70B Turbo

Window TTFT ms — mean · p50 · p90 Throughput tok/s — mean · p50 · p90
May 2026 ~2660 ~2330 ~4440 ~12 ~13 ~15
June 2026 ~2470 ~2290 ~4400 ~12 ~13 ~14
July 2026 ~2590 ~2240 ~3570 ~13 ~13 ~15

GLM 5.2 (FP4)

Window TTFT ms — mean · p50 · p90 Throughput tok/s — mean · p50 · p90
May 2026 ~1490 ~1380 ~2480 ~55 ~58 ~69
June 2026 ~1600 ~1390 ~2760 ~56 ~57 ~61
July 2026 ~1680 ~1510 ~2420 ~49 ~53 ~61

DeepSeek V4 Flash

Window TTFT ms — mean · p50 · p90 Throughput tok/s — mean · p50 · p90
May 2026 ~1600 ~1490 ~2790 ~71 ~76 ~81
June 2026 ~1740 ~1540 ~3050 ~69 ~73 ~78
July 2026 ~1640 ~1490 ~2810 ~72 ~75 ~89

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 DeepInfra yet — the ledger above opens as soon as the first practitioner submission clears verification.

Key Facts

The record, in brief

CategoryAggregator
HeadquartersPalo Alto, USA
Founded2022
Model accessOpen-weight
Flagship / referencePay-per-token open models · Llama 3.3 70B
OpenAI-compatible APIYes
Indicative price (in / out) $0.35 / 1M tokens / $0.40 / 1M tokens medium
Reference throughput~13 tok/s
Reference TTFT~2250 ms
Stated uptime99.5%
ComplianceSOC 2 Type II · GDPR
Composite reputation 81 / 100 · ≈ 4.1 / 5 · 305 notes

Reviewer Notes

What the field says

The 305 notes behind DeepInfra'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: Aggressive per-token pricing on popular open models. Detractors return to another: Fewer enterprise features than larger platforms.

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.

Cross-references

Related dossiers · see also

Peer entries in the same or adjacent category, for comparison against DeepInfra's standing.

Back to the Reputation Index Read §01 — The Landscape