Evidaxis
Open data · CC0 · weekly

Evidaxis

A free weekly CC0 archive of measured facts about open AI systems: commits per week, citations to date and verified dependents, each with its source and date.

5,914 systems in this snapshot· 9 niches + not yet assigned (5,778 systems)· snapshot week 39 of 2026 (2026-w39), dated 2026-09-26

Niches

Niche name, A to Z. Each niche keeps its own median; medals and Rising are computed inside a niche.

Agent Frameworks

17 systems

Median commits/wk
≈21.5
Median citations
60 (n = 5)
Median direct dependents
151 (n = 17)
At 2× the niche median or more
5

Rising: AgentScope

AI Drug Discovery

19 systems

Median commits/wk
0
Median citations
433 (n = 19)
Median direct dependents
5 (n = 2)
At 2× the niche median or more
7

Rising: none

Coding Agents

20 systems

Median commits/wk
≈28.15
Median citations
30 (n = 5)
Median direct dependents
21 (n = 15)
At 2× the niche median or more
3

Rising: Qwen Code

GUI & Computer-Use Agents

7 systems

Median commits/wk
≈0.2
Median citations
≈3.5 (n = 6)
Median direct dependents
84 (n = 2)
At 2× the niche median or more
2

Rising: none

Inference & Serving

13 systems

Median commits/wk
32
Median citations
13 (n = 5)
Median direct dependents
8 (n = 8)
At 2× the niche median or more
3

Rising: none

Media Generation

15 systems

Median commits/wk
0
Median citations
7 (n = 6)
Median direct dependents
119 (n = 3)
At 2× the niche median or more
1

Rising: none

Multimodal Foundation Models

13 systems

Median commits/wk
0
Median citations
26 (n = 7)
Median direct dependents
≈33.5 (n = 2)
At 2× the niche median or more
1

Rising: none

Post-training & Alignment

16 systems

Median commits/wk
≈11.95
Median citations
11 (n = 4)
Median direct dependents
3 (n = 8)
At 2× the niche median or more
2

Rising: none

Robotics & Embodied AI

16 systems

Median commits/wk
≈0.65
Median citations
45 (n = 7)
Median direct dependents
81 (n = 7)
At 2× the niche median or more
4

Rising: Isaac Lab

Medals this week: 3 gold, 9 silver, 16 bronze · Rising: AgentScope, Isaac Lab, Qwen Code

What the figures mean: Commits per week · Citation · Verified direct dependent · Niche medal · Rising

How a figure is computed

Three axes. One gate: any two converge.

We hold a longitudinal record of open AI momentum: every system in the census, every week, each input hash-pinned with raw provenance published. Size is never scored. We measure the rate of change of public signals, and publish Rising only when two independent axes converge. Open data, CC0.

Development, citations and direct dependents.

Each axis is normalized within its cohort, so “rising” means rising relative to its peers. Stars and raw size are never scored. The gate publishes recognition only, and recognition only ever points up.

velocity

Development velocity

The log-slope of commit activity, a fast, leading signal of where work is accelerating.

citation

Citation momentum

The log-slope of yearly citations, a slow, confirming signal that the work is taking hold.

dependents

Direct-dependents momentum (deps.dev)

Confirmed-clean weekly package-union counts, measured as a size-adjusted log-slope within the frozen panel.

gate

The convergence gate

A system receives a Rising badge only when any two of three measurable axes converge. History, axis coverage, cohort floor, and reference-measurement rules are evaluated mechanically each week.

Why convergence? At least two axes must independently meet their rising rules. The third axis expands coverage; thresholds stay fixed. This axis measures the frozen-panel trajectory of the package set selected and linkage-verified as of 2026-07-21. Systems outside that panel are not measured on this axis. Residual survivorship beyond the frozen universe is disclosed, not denied.Full chart wall on the coverage atlas →

Convergence is rare, and that is the point.

The number that matters here is coverage, not the Rising column. We hold 5914 systems in period 2026-w39, measured every week, each input hash-pinned. Published Rising signals: 3. Every weekly snapshot extends the observation history used by the next evaluation.

Risingany two axes
3
published positive signals
History sufficient≥4 weeks
188
distinct weekly observations
At least two of three axes measurablecoverage
62
5 with all three axes

Catalog

Systems with an assigned niche in snapshot 9d2fe956adbc, in niche and name order.

Catalog · snapshot
Catalog: systems in this snapshot (cohort then alphabetical; optional client-side momentum sort)
System Cohort Velocity z Citation z Direct-dependents z Momentum Status
Cosmos Robotics & Embodied AI +2.75 no axis no axis 84.4 1 of 3 axes measurable
dora-rs Robotics & Embodied AI +3.20 no axis -2.75 52.9 2 of 3 axes measurable
Genesis Robotics & Embodied AI +0.12 no axis no axis 51.5 1 of 3 axes measurable
Gymnasium Robotics & Embodied AI +1.80 no axis -0.20 60.1 2 of 3 axes measurable
Habitat-Lab Robotics & Embodied AI -0.03 -0.30 no axis 48 2 of 3 axes measurable
Isaac Lab Robotics & Embodied AI +2.90 +1.37 no axis 76.7 Rising
Isaac-GR00T Robotics & Embodied AI -2.00 no axis no axis 25 1 of 3 axes measurable
LeRobot Robotics & Embodied AI -0.34 no axis +0.00 47.9 2 of 3 axes measurable
LIBERO Robotics & Embodied AI +0.42 no axis no axis 55.3 1 of 3 axes measurable
ManiSkill Robotics & Embodied AI -1.57 -0.69 no axis 35.9 2 of 3 axes measurable
MuJoCo Robotics & Embodied AI +0.53 -0.37 +1.28 56 3 of 3 axes measurable
openpi Robotics & Embodied AI -0.91 no axis no axis 38.6 1 of 3 axes measurable
OpenVLA Robotics & Embodied AI -0.03 no axis no axis 49.6 1 of 3 axes measurable
RDT-1B Robotics & Embodied AI +0.53 no axis no axis 56.6 1 of 3 axes measurable
RoboCasa Robotics & Embodied AI +0.51 no axis no axis 56.4 1 of 3 axes measurable
robosuite Robotics & Embodied AI +0.25 no axis +0.67 55.8 2 of 3 axes measurable
AlphaFold3 AI Drug Discovery -0.23 +0.40 no axis 51.1 2 of 3 axes measurable
BioEmu AI Drug Discovery +2.78 no axis no axis 84.8 1 of 3 axes measurable
Boltz AI Drug Discovery -2.76 +2.04 no axis 45.5 2 of 3 axes measurable
Chai-1 AI Drug Discovery -2.98 no axis no axis 12.7 1 of 3 axes measurable
DeepChem AI Drug Discovery +3.38 -0.73 +0.00 61 3 of 3 axes measurable
DiffDock AI Drug Discovery -0.05 -0.35 no axis 47.5 2 of 3 axes measurable
DiffSBDD AI Drug Discovery -0.36 no axis no axis 45.4 1 of 3 axes measurable
ESM / ESMFold AI Drug Discovery +0.23 +0.48 no axis 54.5 2 of 3 axes measurable
Foldseek AI Drug Discovery +2.89 +0.65 no axis 72.2 2 of 3 axes measurable
gnina AI Drug Discovery -1.23 -0.36 no axis 40.1 2 of 3 axes measurable
Graphormer AI Drug Discovery +3.08 -0.23 no axis 67.8 2 of 3 axes measurable
LigandMPNN AI Drug Discovery -0.31 no axis no axis 46.1 1 of 3 axes measurable
OpenFold AI Drug Discovery +0.18 -0.42 no axis 48.5 2 of 3 axes measurable
ProteinMPNN AI Drug Discovery -0.00 -0.51 no axis 46.8 2 of 3 axes measurable
Protenix AI Drug Discovery -2.97 no axis no axis 12.9 1 of 3 axes measurable
RFdiffusion AI Drug Discovery -2.86 +0.07 no axis 32.6 2 of 3 axes measurable
SaProt AI Drug Discovery -0.31 +1.10 no axis 54.9 2 of 3 axes measurable
TorchDrug AI Drug Discovery -0.05 -0.17 no axis 48.6 2 of 3 axes measurable
Uni-Mol AI Drug Discovery -0.14 +0.52 no axis 52.4 2 of 3 axes measurable
Agentless Coding Agents -0.97 no axis no axis 37.9 1 of 3 axes measurable
Aider Coding Agents -0.52 no axis -0.37 44.4 2 of 3 axes measurable
ChatDev Coding Agents -1.17 no axis no axis 35.4 1 of 3 axes measurable
Cline Coding Agents +1.30 no axis -0.70 53.8 2 of 3 axes measurable
Codex CLI Coding Agents +1.87 no axis no axis 73.4 1 of 3 axes measurable
Continue Coding Agents -0.73 no axis +0.00 45.5 2 of 3 axes measurable
Crush Coding Agents -0.12 no axis no axis 48.5 1 of 3 axes measurable
Gemini CLI Coding Agents -1.59 no axis -0.38 37.7 2 of 3 axes measurable
Goose Coding Agents +0.74 no axis no axis 59.3 1 of 3 axes measurable
Kilo Code Coding Agents -0.20 no axis +1.82 60.1 2 of 3 axes measurable
mini-swe-agent Coding Agents -0.99 no axis +0.27 45.5 2 of 3 axes measurable
Open Interpreter Coding Agents -0.76 no axis +2.11 58.5 2 of 3 axes measurable
OpenCode Coding Agents -0.96 no axis +1.28 52 2 of 3 axes measurable
OpenHands Coding Agents +0.14 no axis no axis 51.7 1 of 3 axes measurable
Qwen Code Coding Agents +2.67 no axis +1.60 76.7 Rising
SWE-agent Coding Agents +0.18 no axis no axis 52.3 1 of 3 axes measurable
SWE-bench Coding Agents +1.69 no axis -0.20 59.3 2 of 3 axes measurable
Tabby Coding Agents +0.61 no axis no axis 57.6 1 of 3 axes measurable
Void Coding Agents +0.45 no axis -0.28 51 2 of 3 axes measurable
Zed Coding Agents +0.16 no axis -0.67 46.8 2 of 3 axes measurable
AIBrix Inference & Serving +0.71 no axis no axis 58.8 1 of 3 axes measurable
KTransformers Inference & Serving -1.54 no axis no axis 30.7 1 of 3 axes measurable
llama.cpp Inference & Serving -0.17 no axis +0.22 50.3 2 of 3 axes measurable
llm-d Inference & Serving -1.08 no axis no axis 36.5 1 of 3 axes measurable
LMCache Inference & Serving +0.01 no axis no axis 50.1 1 of 3 axes measurable
LMDeploy Inference & Serving +0.00 no axis no axis 50 1 of 3 axes measurable
Mooncake Inference & Serving +1.59 no axis no axis 69.9 1 of 3 axes measurable
Nano-vLLM Inference & Serving -2.04 no axis no axis 24.5 1 of 3 axes measurable
Ollama Inference & Serving -0.08 no axis +0.00 49.5 2 of 3 axes measurable
SGLang Inference & Serving +0.53 no axis -0.67 49.1 2 of 3 axes measurable
TensorRT-LLM Inference & Serving +0.48 no axis no axis 56 1 of 3 axes measurable
vLLM Inference & Serving +1.03 no axis -3.00 37.7 2 of 3 axes measurable
Xinference Inference & Serving +0.92 no axis no axis 61.5 1 of 3 axes measurable
AgentScope Agent Frameworks +2.83 no axis +1.95 79.8 Rising
Agno Agent Frameworks -0.17 no axis -0.54 45.6 2 of 3 axes measurable
AutoGen Agent Frameworks +0.29 no axis -0.63 47.9 2 of 3 axes measurable
CAMEL Agent Frameworks -0.63 +0.00 +0.76 50.6 3 of 3 axes measurable
CrewAI Agent Frameworks -0.21 no axis +0.89 54.3 2 of 3 axes measurable
DSPy Agent Frameworks +1.18 no axis -0.40 54.9 2 of 3 axes measurable
Google ADK Agent Frameworks +1.51 no axis +0.36 61.7 2 of 3 axes measurable
LangGraph Agent Frameworks -1.76 no axis +0.29 40.8 2 of 3 axes measurable
Letta Agent Frameworks -0.18 -0.67 -1.43 40.5 3 of 3 axes measurable
Mastra Agent Frameworks +1.16 no axis +0.00 57.3 2 of 3 axes measurable
MetaGPT Agent Frameworks +0.54 +1.33 -0.96 53.8 3 of 3 axes measurable
Microsoft Agent Framework Agent Frameworks +0.28 no axis +3.00 70.5 2 of 3 axes measurable
OpenAI Agents SDK Agent Frameworks +1.34 no axis -0.23 57 2 of 3 axes measurable
PydanticAI Agent Frameworks +1.70 no axis +0.23 62.1 2 of 3 axes measurable
Qwen-Agent Agent Frameworks -0.30 no axis -0.67 43.9 2 of 3 axes measurable
smolagents Agent Frameworks -0.65 no axis -0.84 40.7 2 of 3 axes measurable
Strands Agents Agent Frameworks -0.16 no axis +0.89 54.6 2 of 3 axes measurable
AReaL Post-training & Alignment -0.26 no axis no axis 46.7 1 of 3 axes measurable
ART Post-training & Alignment +2.11 no axis no axis 76.4 1 of 3 axes measurable
Axolotl Post-training & Alignment +0.20 no axis no axis 52.5 1 of 3 axes measurable
LLaMA-Factory Post-training & Alignment -1.67 no axis no axis 29.2 1 of 3 axes measurable
ms-swift Post-training & Alignment +0.19 no axis no axis 52.4 1 of 3 axes measurable
NeMo-RL Post-training & Alignment +1.97 no axis no axis 74.6 1 of 3 axes measurable
Open R1 Post-training & Alignment -0.60 no axis no axis 42.5 1 of 3 axes measurable
Open-Instruct Post-training & Alignment -1.53 no axis no axis 30.9 1 of 3 axes measurable
OpenRLHF Post-training & Alignment -0.71 no axis no axis 41.1 1 of 3 axes measurable
prime-rl Post-training & Alignment +0.96 no axis no axis 62 1 of 3 axes measurable
ROLL Post-training & Alignment -0.89 no axis no axis 38.9 1 of 3 axes measurable
Search-R1 Post-training & Alignment +0.34 no axis no axis 54.2 1 of 3 axes measurable
Trinity-RFT Post-training & Alignment -0.81 no axis no axis 39.9 1 of 3 axes measurable
TRL Post-training & Alignment +0.40 no axis +0.00 52.5 2 of 3 axes measurable
Unsloth Post-training & Alignment +2.01 no axis +0.00 62.6 2 of 3 axes measurable
verl Post-training & Alignment -0.77 no axis no axis 40.4 1 of 3 axes measurable
ACE-Step Media Generation +0.16 no axis no axis 52 1 of 3 axes measurable
ComfyUI Media Generation +2.69 no axis no axis 83.7 1 of 3 axes measurable
Diffusers Media Generation +2.58 no axis -0.67 61.9 2 of 3 axes measurable
Fish Speech Media Generation -3.14 no axis no axis 10.7 1 of 3 axes measurable
FLUX.1 Media Generation -0.11 no axis no axis 48.7 1 of 3 axes measurable
HunyuanVideo Media Generation +0.68 no axis no axis 58.5 1 of 3 axes measurable
Kokoro-82M Media Generation +0.06 no axis +0.00 50.4 2 of 3 axes measurable
LTX-Video Media Generation +0.03 no axis no axis 50.4 1 of 3 axes measurable
Open-Sora Media Generation -3.13 no axis no axis 10.9 1 of 3 axes measurable
Qwen-Image Media Generation +0.07 no axis no axis 50.9 1 of 3 axes measurable
Sana Media Generation +3.06 no axis no axis 88.3 1 of 3 axes measurable
SkyReels-V2 Media Generation +0.09 no axis no axis 51.1 1 of 3 axes measurable
Stable Diffusion 3.5 Media Generation +0.34 no axis no axis 54.3 1 of 3 axes measurable
Wan2.2 Media Generation +2.96 no axis no axis 87 1 of 3 axes measurable
YuE Media Generation +3.04 no axis no axis 88 1 of 3 axes measurable
DeepSeek-OCR Multimodal Foundation Models +0.00 no axis no axis 50 1 of 3 axes measurable
DeepSeek-VL2 Multimodal Foundation Models +0.00 no axis no axis 50 1 of 3 axes measurable
Eagle 2.5 Multimodal Foundation Models -3.00 no axis no axis 12.5 1 of 3 axes measurable
FastVLM Multimodal Foundation Models +2.83 no axis no axis 85.4 1 of 3 axes measurable
GLM-V Multimodal Foundation Models -3.00 no axis no axis 12.5 1 of 3 axes measurable
InternVL Multimodal Foundation Models +0.00 no axis no axis 50 1 of 3 axes measurable
Kimi-VL Multimodal Foundation Models +0.00 no axis no axis 50 1 of 3 axes measurable
LLaVA Multimodal Foundation Models +0.00 no axis no axis 50 1 of 3 axes measurable
MiniCPM-V Multimodal Foundation Models -3.00 no axis no axis 12.5 1 of 3 axes measurable
moondream Multimodal Foundation Models -2.56 no axis no axis 18 1 of 3 axes measurable
Ovis2.5 Multimodal Foundation Models +0.68 no axis no axis 58.4 1 of 3 axes measurable
Qwen3-VL Multimodal Foundation Models +0.00 no axis +0.00 50 2 of 3 axes measurable
SmolVLM Multimodal Foundation Models -1.21 no axis no axis 34.8 1 of 3 axes measurable
Agent S GUI & Computer-Use Agents +0.34 no axis no axis 54.3 1 of 3 axes measurable
Agent-E GUI & Computer-Use Agents +0.75 no axis no axis 59.3 1 of 3 axes measurable
Browser Use GUI & Computer-Use Agents -1.55 no axis -0.67 36.1 2 of 3 axes measurable
OmniParser GUI & Computer-Use Agents -0.17 no axis no axis 47.9 1 of 3 axes measurable
OSWorld GUI & Computer-Use Agents -2.58 no axis no axis 17.7 1 of 3 axes measurable
ShowUI GUI & Computer-Use Agents -0.39 no axis no axis 45.1 1 of 3 axes measurable
UI-TARS Desktop GUI & Computer-Use Agents +1.30 no axis +0.00 58.1 2 of 3 axes measurable
Default order: cohort, then alphabetical. Momentum maps the mean of present-axis z-scores to 0-100. Cohort-normalized z-scores · measurement states are derived from source coverage and weekly history · "no axis" = that axis is not measurable. Sort-by-momentum is optional client-side only. Two- and three-axis momentum scores are not strictly comparable.

5778 more systems are measured weekly and await a niche