model
Qwen3 30B A3B Instruct 2507 llm
A model record: onboarded once for the project and used by any eval's arms.
- weights
- open
- licence
- Apache-2.0
- aliases
qwen3:30b-a3b-instruct-2507-q4_K_MNVFP4/Qwen3-30B-A3B-Instruct-2507-FP4eval-qwen3-30b-a3b- settings
- defaults
- onboarding
- onboarded, no arm smoked 2026-10-31
- best result per goal
- #1 of 55 Triage decisions on Orrery's error signals
01deployments and their onboarding
x3 · x3 · ollama 0.40.13 · qwen3:30b-a3b-instruct-2507-q4_K_M · onboarding pass 2026-10-31 · 70.7 tokens/s
- where
- x3
- server or device
- ollama 0.40.13
- build
qwen3:30b-a3b-instruct-2507-q4_K_M19e422b0231392335cfc49cfd172de7034bb1aeabb08aa307cce745c60b272feQ4_K_M- matches the record
- yes
- checked
- 2026-10-31
- onboarded on
- x3
| probe | measurement | flag |
|---|---|---|
| endpoint | provider openai · base url http://127.0.0.1:11434/v1 · model qwen3:30b-a3b-instruct-2507-q4_K_M | — |
| digest found | 19e422b0231392335cfc49cfd172de7034bb1aeabb08aa307cce745c60b272fe | — |
| default | s 6.28 · tokens out 23 · reasoned: no · finish stop | — |
| chosen | s 0.33 · tokens out 23 · reasoned: no · finish stop · settings () | — |
| tokens per s | 70.7 | — |
| deterministic | yes | — |
| json valid | yes | — |
dgx · dgx · ollama 0.40.13 · qwen3:30b-a3b-instruct-2507-q4_K_M · onboarding pass 2026-10-31 · 80.1 tokens/s
- where
- dgx
- server or device
- ollama 0.40.13
- build
qwen3:30b-a3b-instruct-2507-q4_K_M19e422b0231392335cfc49cfd172de7034bb1aeabb08aa307cce745c60b272fe- matches the record
- yes
- checked
- 2026-10-31
- onboarded on
- dgx
| probe | measurement | flag |
|---|---|---|
| endpoint | provider openai · base url http://100.69.49.126:11434/v1 · model qwen3:30b-a3b-instruct-2507-q4_K_M | — |
| digest found | 19e422b0231392335cfc49cfd172de7034bb1aeabb08aa307cce745c60b272fe | — |
| default | s 5.58 · tokens out 23 · reasoned: no · finish stop | — |
| chosen | s 0.29 · tokens out 23 · reasoned: no · finish stop · settings () | — |
| tokens per s | 80.1 | — |
| deterministic | yes | — |
| json valid | yes | — |
mac · mac · ollama 0.40.13 · qwen3:30b-a3b-instruct-2507-q4_K_M · onboarding pass 2026-10-31 · 56.5 tokens/s
- where
- mac
- server or device
- ollama 0.40.13
- build
qwen3:30b-a3b-instruct-2507-q4_K_M19e422b0231392335cfc49cfd172de7034bb1aeabb08aa307cce745c60b272feQ4_K_M- matches the record
- yes
- checked
- 2026-10-31
- onboarded on
- mac
| probe | measurement | flag |
|---|---|---|
| endpoint | provider openai · base url http://127.0.0.1:11435/v1 · model qwen3:30b-a3b-instruct-2507-q4_K_M | — |
| digest found | 19e422b0231392335cfc49cfd172de7034bb1aeabb08aa307cce745c60b272fe | — |
| default | s 16.47 · tokens out 23 · reasoned: no · finish stop | — |
| chosen | s 0.41 · tokens out 23 · reasoned: no · finish stop · settings () | — |
| tokens per s | 56.5 | — |
| deterministic | yes | — |
| json valid | yes | — |
dgx-vllm · dgx · vllm (production's, :8003) · NVFP4/Qwen3-30B-A3B-Instruct-2507-FP4 · onboarding pass 2026-10-31 · 65.0 tokens/s
- where
- dgx
- server or device
- vllm (production's, :8003)
- build
NVFP4/Qwen3-30B-A3B-Instruct-2507-FP4- matches the record
- yes
- checked
- 2026-10-31
- onboarded on
- dgx
| probe | measurement | flag |
|---|---|---|
| endpoint | provider openai · base url http://100.69.49.126:8003/v1 · model NVFP4/Qwen3-30B-A3B-Instruct-2507-FP4 | — |
| digest found | None | — |
| default | s 0.83 · tokens out 23 · reasoned: no · finish stop | — |
| chosen | s 0.35 · tokens out 23 · reasoned: no · finish stop · settings () | — |
| tokens per s | 65.0 | — |
| deterministic | yes | — |
| json valid | yes | — |
hosted · homelab LiteLLM (the fleet's gateway) · eval-qwen3-30b-a3b · onboarding pass 2026-10-31 · 53.4 tokens/s
- where
- homelab LiteLLM (the fleet's gateway)
- build
eval-qwen3-30b-a3b- matches the record
- yes
- checked
- 2026-10-31
- onboarded on
- homelab LiteLLM (the fleet's gateway)
| probe | measurement | flag |
|---|---|---|
| endpoint | provider litellm · base url http://homelab:4001/v1 · model eval-qwen3-30b-a3b | — |
| digest found | None | — |
| default | s 2.18 · tokens out 22 · reasoned: no · finish stop | — |
| chosen | s 0.47 · tokens out 25 · reasoned: no · finish stop · settings () | — |
| tokens per s | 53.4 | — |
| deterministic | no | — |
| json valid | yes | — |
What it can do, across its deployments: reasons: no · json mode: yes
02settings history
No change of settings or build is recorded.
03arms
| arm | deployment | goal | its settings | arm smoke | result |
|---|---|---|---|---|---|
| triage_hosted_qwen3_30b_a3b_v1 · rescored 5f2c057c orrery_triage_v1_labelled | hosted | Triage decisions on Orrery's error signals | — | none | triage_cost 0.708 |
| triage_local_dgx_qwen3_30b_a3b_v1 · rescored 5f2c057c orrery_triage_v1_labelled | dgx | Triage decisions on Orrery's error signals | call_timeout_s 600 | none | triage_cost 0.892 |
| triage_local_dgx_vllm_qwen3_30b_a3b_nvfp4_v1 · rescored 5f2c057c orrery_triage_v1_labelled | dgx | Triage decisions on Orrery's error signals | call_timeout_s 600 | none | triage_cost 0.400 |
| triage_local_mac_qwen3_30b_a3b_v1 · rescored 5f2c057c orrery_triage_v1_labelled | mac | Triage decisions on Orrery's error signals | call_timeout_s 600 | none | triage_cost 0.831 |
| triage_local_x3_qwen3_30b_a3b_v1 · rescored 5f2c057c orrery_triage_v1_labelled | x3 | Triage decisions on Orrery's error signals | call_timeout_s 600 | none | triage_cost 0.723 |