Levels of evidence
Published by the source and re-read in the original document or database.
Recomputed by us from microdata and matching a figure the source itself publishes.
Our own tabulation, reproducible and tested, with no published counterpart to check against.
What “open-weight” means here
A model counts as open-weight only when automated checks found its weight files in a public repository that can be read without signing in, and only from the first documented date of that availability; where the first archived copy of a repository shows a page without the weight files, the date is the day the files appeared. The licence is a separate question: a verified permissive licence, a non-commercial or unverified licence, and gated access with unread terms are kept apart. No model is confirmed as closed-weight merely because its weights could not be found, on OpenRouter or on Hugging Face; the rule for confirming closed weights is being revalidated, and the nine decisions recorded so far are suspended until it is. Openness is never inferred from how a model is served.
Data as of
The report is titled 2026, but every source has its own date. The chapters give each figure’s date beside it; this table puts them in one place.
| Source | What it measures | Period covered | Read or retrieved | Chapters |
|---|---|---|---|---|
| Hugging Face Hub counters | 30-day downloads of the tracked models; downloads of quantised copies | August 2024 to October 2026, read on the 1st of each month | 22 October 2026 (snapshots) | 01, 03, 06 |
| ATOM Project tracked-models list | Which models are tracked, and their developers’ regions | List at commit d4b1709 | 20 October 2026 | 01 |
| OpenAlex | Works naming a model family | 2023 to 2025 | 26 October 2026 | 01 |
| Epoch AI benchmark runs | Scores on three tests | Models released by 10 November 2026 | Archive of 10 November 2026 | 02 |
| Artificial Analysis, VIALS-AA | Scientific images interpreted, all three attempts right (permission requested) | Models released by 10 November 2026 | 11 November 2026 (as displayed) | 02 |
| Vals AI, Terminal-Bench 4.0 | Long agentic tasks completed, with cost and time per task, the same agent for every model | Models released by 10 November 2026 | Leaderboard of 9 November 2026 | 02, 04 |
| DeepSWE (Datacurve), via Epoch AI | Programming tasks solved, the same agent for every model | Models released to early September 2026 | Archive of 7 November 2026; Epoch AI’s review of DeepSWE 1.1, 7 September 2026 | 02 |
| LM Arena | How it labels models (cited, not scored) | Snapshots to 15 October 2026 | 19 October 2026 | 02 |
| OpenRouter rankings | Daily tokens by model version | 1 January 2025 to 14 October 2026 | 15 October 2026 | 01, 03 |
| OpenRouter, requests sorted by subject | Tokens by subject, the most used models | Weeks of 21 September to 2 November 2026 | 9 November 2026 | 03 |
| OpenRouter providers and endpoints | Monthly tokens by provider; endpoints’ prices and declared precision | The month to 9 November 2026 | 9 November 2026 (as displayed) | 06 |
| Cloudflare Radar, Workers AI | Shares of requests by model and by task | June 2025 to October 2026 | 4 November 2026 | 01, 03 |
| Providers’ price pages and GPU list prices | API prices per token; GPU rental prices per hour | November 2026 | November 2026 | 06 |
| SemiAnalysis InferenceX | Measured throughput per GPU on given configurations | Configurations as published | 9 November 2026 | 06 |
| Menlo Ventures, Ramp AI Index, Vercel AI Gateway, ETR/Futurum, Futurum | Published figures on companies’ use and spending | 2025 to September 2026 | Cited as published; dates in chapter 05 | 05 |
| JetBrains and Stack Overflow developer surveys | Microdata on developers’ use of models | 2025 survey waves | Dates in chapters 05 and 06 | 05, 06 |
| Eurostat, isoc_eb_ai | Enterprises using AI | Surveys of 2021 to 2025 | 19 October 2026 | 05 |
| Developers’ model cards and licences | What each named release is | Releases to November 2026 | 7 November 2026 | 02, 06 |
Figures held back
Some figures exist in the project but are not placed in this edition:
| Figure | Subject | Why not shown |
|---|---|---|
| 5.4 | Open versus closed models reaching production, by organisation size (developers) | Not placed: no web chart was ever built. |
| 6.3 | Where surveyed organisations host generative AI: LF 2024 and CNCF 2025, side by side | Not placed in this edition: held as an internal placeholder for its side-by-side design (Sprint 09 brief). The datasets state CDLA-Permissive-2.0 in their introductions and on their Zenodo records. |
| 3.4 | Open-weight share of observed tokens, lower bound (2026-Q3) | Internal until the high-impact classification review is completed and the classification-date sensitivity is documented. The chart is held back; chapter 03 cites its third-quarter lower bound in the text, as DEC-076 allows. |
Our interest, and how we handle it
Helmcode, which publishes this report, builds and operates private inference infrastructure for open-weight models. It therefore has a commercial interest in the subject of the report, and most directly in what running open models costs. These safeguards apply throughout:
- Every number is linked to its source and to the file that produces it; the figures are drawn by the same code that writes these pages, so each can be rebuilt.
- Independent observations (benchmark runs, surveys, public price lists, usage rankings) are kept apart from claims made by companies, which are cited as claims.
- No Helmcode figure is used in any calculation.
- Where a result depends on assumptions, the assumptions are shown across their full range and the cases that cut against us are reported alongside the others.
- Any comparison of dedicated infrastructure with API prices was to be read by a reviewer outside Helmcode before it was published (decision DEC-110). That review has not taken place. On 9 November 2026 the editor lifted the requirement for chapter 06, which publishes the comparison as a modelled scenario, not as an observed cost.
This report has not been peer-reviewed.
Sources and rights
No partner, sponsor or institutional endorsement is implied.
Data: Epoch AI, OpenRouter, Eurostat, JetBrains, LM Arena (CC BY 4.0); Linux Foundation Research and CNCF microdata (CDLA-Permissive-2.0); Hugging Face Hub counters (used with Hugging Face’s permission); Cloudflare Radar AI Insights, Workers AI model data (used with Cloudflare’s permission); OpenAlex (CC0); OpenHands Index (Apache-2.0). Each figure states its own source, base and evidence level. Type: Newsreader, IBM Plex Sans, IBM Plex Sans Condensed, Roboto Mono (SIL OFL 1.1).
- Epoch AI, Capabilities & benchmarking and Data on AI models (CC BY 4.0).
- OpenRouter rankings-daily (CC BY 4.0): source line “OpenRouter (openrouter.ai/rankings), as of …” on every figure.
- Eurostat, isoc_eb_ai (CC BY 4.0).
- JetBrains Developer Ecosystem Survey 2025 raw data (CC BY 4.0).
- LM Arena leaderboard dataset (CC BY 4.0).
- Linux Foundation Research and CNCF microdata on Zenodo: CDLA-Permissive-2.0, as stated in the dataset introductions and on the Zenodo records.
- Stack Overflow Developer Survey 2025 (Open Database License, ODbL 1.0).
- Commercial surveys marked all rights reserved are cited as text only; none of their charts is reproduced.
Licences: CC BY 4.0 · CDLA-Permissive-2.0 · CC0 1.0 · Apache-2.0 · MIT · ODbL 1.0.
Reproducibility
Every number on these pages is read from the project's analysis files by the same code that builds the page; the charts are drawn in the browser from the embedded data, which can be inspected under each figure.
