[mle]/DaBench
Leaderboard

Beam 1.0 501B A23B

reflection/beam-1.0-501b-a23b

Hostedreflection274 tok/sopenrouterserved by NebiusFP81M context
Run recordCompare
Overall score
57

85 of 153 test cases fully passed

Scores by task

Scores give partial credit. Passed counts fully correct answers only. Select a task to read its test cases.

  1. 55Agents8 of 21 passedp90 latency 37.0s
  2. 81Structured Extraction17 of 21 passedp90 latency 19.9s
  3. 48Coding10 of 21 passedp90 latency 34.5s
  4. 69Grounded Answers15 of 21 passedp90 latency 19.9s
  5. 78Instruction Following14 of 21 passedp90 latency 31.5s
  6. 57Business Logic12 of 21 passedp90 latency 20.1s
  7. 51Security & permissions8 of 21 passedp90 latency 34.4s
  8. 17Diagnosis & investigation1 of 6 passedp90 latency 1m 3s

Run it yourself

The API call the benchmark made, and the settings behind this score.

Deployment
Hosted
API
OpenRouter
Serving provider
Nebius
Precision
FP8
Context window
1M
Throughput
274 tok/s
Cost of a suite run
$0.83
Cost per case
$0.0054
Time per case
17.2s
Measured
2026-10-16
Run this buildOpenRouterHosted
curl https://openrouter.ai/api/v1/chat/completions \
  -H "Authorization: Bearer $OPENROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"reflection/beam-1.0-501b-a23b","provider":{"order":["nebius/fp8"],"allow_fallbacks":false},"service_tier":"default","messages":[{"role":"user","content":"Hello"}]}'

The request the benchmark sent, pinned to the provider that served it.

One email when a new model lands on the board or in the showcase. Unsubscribe anytime.

LeaderboardCompare Beam 1.0 501B A23B with another model