Category

Speech: RL environments and AI benchmarks

AI benchmarks
34
Shares entries with
3 categories

Snapshot 2026-10-11. How categories are built · Where the data comes from

In short

What are RL environments for speech?
A simulated workplace where an AI agent practises a task (a code repository, a browser, business software) and an automatic grader scores each attempt. Labs train on thousands of these attempts. Bought to improve models.

Only benchmarks use the Speech tag: the RL environment directories file their vendors under other categories. Browse all RL environments.

AI benchmarks 34

None confirmed yet.

34 not yet confirmedListed in the catalog; no line from its own site or paper confirms it yet.
  1. AudioMC - Text Output
  2. IHBench
  3. PolySpeech-100
  4. VoiceBench
  5. AudioMC - Audio Output
  6. SpeechEQ
  7. AfriVox-v2
  8. ASR Code-Switch Benchmark
  9. From Text to Voice
  10. KVoiceBench, KOpenAudioBench, KMMAU
  11. PashtoTTS-Bench
  12. When Multiple Scripts Matter
  13. AudioAgentBench
  14. VoiceAssistant-Eval
  15. Dynamic-SUPERB
  16. VoiceAgentBench
  17. FLEURS
  18. OpenSTBench
  19. PRiSM: Benchmarking Phone Realization in Speech Models
  20. Vaani-Benchmark-V1.0
  21. SURE
  22. AudioMC
  23. Benchmarking Large Language Models for Grapheme-to-Phoneme Conversion
  24. Sarashina2.2-TTS
  25. ZONOS2 Technical Report
  26. Speech-MASSIVE
  27. SUPERB
  28. TurnBench
  29. Moroccan Darija ASR Leaderboard
  30. Open Universal Arabic ASR Leaderboard
  31. Benchmarking von ASR-Modellen im deutschen medizinischen Kontext
  32. Beyond Content
  33. CoVoST2
  34. PAREDA

Domains that overlap with Speech

Next: pick another domain. Tags come from the RL Research daily scrape, canonicalized nightly.