Openbenchmarks
Independent benchmarks for build-vs-buy decisions.
- 0 community votes
- 40 DR by Ahrefs +1 (up) since last check
- 2026 founded
We checked the Ahrefs Domain Rating 16 Nov 2026, it changed 6 Nov 2026.
Openbenchmarks measures AI agent tooling across inference APIs, web search, company data, voice agent latency, and speech models through independent
The story

openbenchmarks.com, as visitors see it
Openbenchmarks is an independent benchmarking company that develops open benchmarks measuring how well APIs, tools and agents complete specific tasks. Every benchmark is measured first-hand against known answers—never vendor-reported numbers—and results are published with full methodology and public datasets.
The company evaluates products across multiple domains including web search APIs, company enrichment and funding data APIs, voice agent latency, speech-to-text and text-to-speech models, document processing, and inference providers. Each benchmark measures correctness, cost, and latency as separate axes.
Openbenchmarks maintains independence by ensuring no vendor pays for inclusion, ranking or removal. The evaluation harness is open-source, public datasets are published on Hugging Face, and partners can be ranked below other providers on benchmarks they inform. Results are computed on held-out test sets never published before evaluation runs.
Nice things it does
- Web search API benchmarks across factual lookup, hard retrieval, and multi-hop search tasks
- Company enrichment and funding data API evaluation
- Voice agent latency measurement from real phone calls
- Speech-to-text and text-to-speech model benchmarks
- Document processing parser evaluation on redlined contracts
- Inference provider latency and reliability testing
- Company lookalike API benchmarking for account discovery and prospecting
- Public datasets and open-source evaluation harnesses
- Verified ground truth scoring with held-out test sets
Pricing
Pricing model: Free
Follow Openbenchmarks to get an email when they post a milestone, ship a new product or add a demo video.
Openbenchmarks metrics
Here's how openbenchmarks.com looks to Ahrefs (Domain Rating) and Majestic (links), and what changed since we last checked.
- 40 Domain Rating +1 (up) since last check
- 11 Trust Flow no change since last check
- 28 Citation Flow no change since last check
- 167 Referring domains no change since last check
- 863 External backlinks no change since last check
We checked the Ahrefs Domain Rating 16 Nov 2026, it changed 6 Nov 2026. Their big topic: Computers/Data Formats (10). We checked Majestic 1 Nov 2026, it changed 1 Nov 2026. Thanks to Majestic for the link data. Trust Flow, Citation Flow and Topical Trust Flow are trademarks of Majestic-12 Ltd.
A little about Openbenchmarks's domain
Read more about the domain
Public RDAP record, last read 15 Nov 2026, 20:48 UTC.
- Domain age 3 years
- Registered
- Expires
- Registrar GoDaddy.com, LLC
- DNSSEC Not signed
A little about Openbenchmarks since launch
Our after-launch readings for Openbenchmarks, live on Virela Launch since 1 November 2026.
- Domain Rating (Ahrefs)
- 40 +1 (up) since launch
- 39 at launch (1 Nov 2026), 40 on 15 Nov 2026
- Trust Flow (Majestic)
- 11
- First measured 1 Nov 2026
- Referring domains (Majestic)
- 167
- First measured 1 Nov 2026
Day 14 since launch. Next receipt: day 30, 1 Dec 2026. Readings only: they do not prove what caused a change, and many things can. A metric we have no reading for is left out.
Launch tips, free to read
Our free playbook covers what to do before, on and after launch day, plus which launch directories to try.
- The week before
- Go live
- Rest of the week
- Afterwards
People also ask
How does Openbenchmarks ensure independence?
Openbenchmarks maintains independence by ensuring no vendor pays for inclusion, ranking, or removal. Published rankings come only from a held-out test set run just before publication, whose questions are never published and answer keys never leave Openbenchmarks. Evaluation runs are not identifiable to providers, and no one sees rankings before they go live.
What benchmarks does Openbenchmarks offer?
Openbenchmarks benchmarks web search APIs, company enrichment and funding data APIs, voice agent latency, speech-to-text and text-to-speech models, document processing parsers, inference providers, company lookalike APIs, and multi-turn company search systems. Each benchmark measures correctness, cost, and latency separately.
Are the benchmarks and code open source?
Yes, the evaluation harness is public on github.com/openbenchmarks-labs including prompts, judge configuration, and scoring code. Public dataset splits are available on Hugging Face, and anyone can propose metrics or contribute by opening a pull request.
How can I report a benchmark error?
If you represent a benchmarked provider and believe a result is wrong, write to founders@openbenchmarks.com with the provider name, dataset split, run timestamp, and evidence. Openbenchmarks will re-run the evaluation and publish the result with updated methodology.
Openbenchmarks has a free listing. The owner can unlock the full profile with a paid launch for $5:
- Founder story (locked)
- Milestones (locked)
- Demo video (locked)
- Dofollow website link (locked)