LocalsOnly · measured2026-09-12
LiveCodeBench
Version 6.0 · 93 tasks. Compare only matched task sets and protocols.
DeepSeek V4 Flash 0731 · UD-IQ3_XXS · DGX Spark · 128GB · llama.cpp
● Tested by LocalsOnlyRecorded effort for this run. Wall time and token counts are not a dollar cost; hardware power and utilization also matter.
Tokens per second · see measurement basis
As recorded on this receipt
Reported category outcomes
Each category keeps its own recorded denominator. Some categories exclude unscored trials. Mean reward, when present, is partial credit.
Run configuration & provenance +
- Context
- 32,768
- Temperature
- 1
- Wall time
- 821.6 minutes
- Weight file
- Provider-managed / not reported
- Revision
- Not reported
- Reasoning
- Thinking enabled
- Agent version
- Not reported
- Output token limit
- 131,072
- Dataset revision
- LiveCodeBench release_v6 codegeneration · same 100 ids as lcb-mac-spark-udq4kxl-20260904a
- Harness host
- Mac mini · codegen+checker
- Model host
- DGX Spark spark-5c2f · :8888
- Input tokens
- 54,967
- Output tokens
- 899,437
- Cached input tokens
- Not reported
- Agent turn / step limit
- Not reported
- Concurrency
- 1
- Weight file size
- Not reported
Throughput measurement: Measurement method not separately reported. This should not be treated as standardized decode-only speed.
lcb-codegen-generate.py
Run ID: measured-dgx-spark-unsloth-ud-iq3_xxs-livecodebench-6.0-2026-09-12Task explorer
Passed tasks first, then other outcomes grouped by recorded issue. Use the filters to inspect a particular outcome.
All 100 recorded tasks, before filters or pagination. 25 of 100 matching tasks shown below. Recorded issues can also occur on scored tasks.
Passed1873 d+
Task ID: 1873_d
Recorded name: 1873_d
Reward: 1
Tokens in / out: 593 / 3,273
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 2808+
Task ID: 2808
Recorded name: 2808
Reward: 1
Tokens in / out: 467 / 2,965
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 2848+
Task ID: 2848
Recorded name: 2848
Reward: 1
Tokens in / out: 307 / 1,786
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 2850+
Task ID: 2850
Recorded name: 2850
Reward: 1
Tokens in / out: 425 / 11,977
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 2857+
Task ID: 2857
Recorded name: 2857
Reward: 1
Tokens in / out: 398 / 4,294
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 2954+
Task ID: 2954
Recorded name: 2954
Reward: 1
Tokens in / out: 576 / 519
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 2955+
Task ID: 2955
Recorded name: 2955
Reward: 1
Tokens in / out: 380 / 344
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 3034+
Task ID: 3034
Recorded name: 3034
Reward: 1
Tokens in / out: 342 / 294
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 3046+
Task ID: 3046
Recorded name: 3046
Reward: 1
Tokens in / out: 412 / 7,495
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 3163+
Task ID: 3163
Recorded name: 3163
Reward: 1
Tokens in / out: 455 / 488
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 3193+
Task ID: 3193
Recorded name: 3193
Reward: 1
Tokens in / out: 534 / 538
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 3194+
Task ID: 3194
Recorded name: 3194
Reward: 1
Tokens in / out: 348 / 182
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 3195+
Task ID: 3195
Recorded name: 3195
Reward: 1
Tokens in / out: 395 / 830
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 3199+
Task ID: 3199
Recorded name: 3199
Reward: 1
Tokens in / out: 372 / 2,057
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 3206+
Task ID: 3206
Recorded name: 3206
Reward: 1
Tokens in / out: 425 / 334
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 3229+
Task ID: 3229
Recorded name: 3229
Reward: 1
Tokens in / out: 717 / 9,348
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 3265+
Task ID: 3265
Recorded name: 3265
Reward: 1
Tokens in / out: 515 / 4,216
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 3299+
Task ID: 3299
Recorded name: 3299
Reward: 1
Tokens in / out: 445 / 23,434
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 3320+
Task ID: 3320
Recorded name: 3320
Reward: 1
Tokens in / out: 390 / 1,837
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 3353+
Task ID: 3353
Recorded name: 3353
Reward: 1
Tokens in / out: 288 / 889
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 3360+
Task ID: 3360
Recorded name: 3360
Reward: 1
Tokens in / out: 454 / 3,952
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 3387+
Task ID: 3387
Recorded name: 3387
Reward: 1
Tokens in / out: 419 / 7,499
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 3406+
Task ID: 3406
Recorded name: 3406
Reward: 1
Tokens in / out: 593 / 6,611
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 3438+
Task ID: 3438
Recorded name: 3438
Reward: 1
Tokens in / out: 584 / 5,520
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.PassedTask ID 3450+
Task ID: 3450
Recorded name: 3450
Reward: 1
Tokens in / out: 554 / 669
Recorded issue: No failure explanation in this task record.
The receipt contains outcome metadata. Full prompts and execution transcripts are not included in this view.