BANTAM FACTORY

BANTAM FACTORY / fight card

Context packet.

Fit complete, attributable context sections into a strict UTF-8 byte budget.

Matchup completed with follow-up runs. DeepSeek Harness, Hermes · 2026-09-08; Pi · 2026-09-08. Original BANTAM FACTORY result retained. Task, starter, grader and local model hashes match. Run conditions ↓

Watch the work Try what it built ↓

BANTAM FACTORY'S LOCAL SETUP

NVIDIA RTX 4090 · 24 GB

Qwen 27B · Q4_K_P · 72,192 context tokens

Recorded generation speed105.2 tok/s14/14 timed requests
What does this mean for my setup?

This run used Qwen 27B on NVIDIA RTX 4090 · 24 GB. Compare generation speed using the same model and quantization. GPU memory, CPU offload, context size and prompt caching also affect performance.

Generation speed is measured while the server produces tokens. The task clock also includes reading context, running tools and testing. A GPU speed difference does not translate directly into the same change in total task time.

Fresh prompt processing: 1,632 tok/s · 14/14 timed requests. Rates use total recorded tokens divided by total server time for each phase. Hardware is operator-confirmed; timings come from the saved server responses.

The recorded factory floor

Meet the contenders.

5 local configurations.
1 frontier reference, shown separately.
Choose any contender. Open its work.

Recorded system comparison

Context packet

Recorded outcomes, elapsed times and token accounting
SystemOutcomeTimeGroupsProject acceptedClean finishInputOutputCached inputFresh inputScope
BANTAM FACTORY · localQwen 27B · same local modelPASS58.1s5/5YesYes109,5864,20693,59015,996Recorded totals
OpenCodeQwen 27B · same local modelOUTPUT_ONLY600.0s5/5YesNo623,29738,632532,79890,499Measured subset
Codex · AstraGPT-6 Astra · native CLIPASS110.6s5/5YesYes75,1962,83562,33612,860Recorded totals
DeepSeek HarnessQwen 27B · same local modelOUTPUT_ONLY600.0s5/5YesNo859,86839,085825,84234,026Measured subset
HermesQwen 27B · same local modelPASS512.2s5/5YesYes356,32433,735330,53925,785Measured subset
PiQwen 27B · same local modelPASS350.7s5/5YesYes415,12725,365408,8796,248Recorded totals

BUILT IN THIS FIGHT. READY TO TRY.

Keep what matters.

BANTAM FACTORY built this context packer in 58.1 seconds. Change the budget. Edit the notes. See what fits.

See packet ↓

The tool runs in your browser.

Loading the tool…

YOUR CONTEXT PACKET

The recorded packing functions run here with browser byte counting. This interface is a demo built around that output; it was not part of the timed task. Your edits stay in this browser. Source & adapter receipts ↗

Method & evidenceThe work order and run conditions.Read the method

The recorded comparison

1 included work order: BUILD — Context packet (repeat 1). 6 systems and 6/6 recorded attempts. Every arm receives the supplied materials for its work order; independent acceptance checks remain separate from worker tests.

Local configurations and frontier references are distinguished in each lane. A frontier result is not a same-model comparison. All included outcomes remain visible; missing results are not zero-cost runs and do not count as passes. Elapsed times and completion status are reported separately.

These are selected development work orders. The original and follow-up recording windows, model settings, tool policies and time limits are documented with each card.

Reading an outcome

PASS: project accepted and clean harness completion. OUTPUT_ONLY: the artifact passed but the run did not reach accepted completion. FAIL: a required check failed. Timeouts and unrecorded tasks remain visible.

Passing independent groups alone does not override a failed public suite or protected-file check. Accepted project and clean finish are separate facts. The independent groups check the supplied work order.

Clocks & counters

Replay aligns each run's recorded start to zero. Runs did not all start simultaneously. Counters use actual saved response times, not animated estimates. Untimed evidence is not assigned an invented timestamp.

Input includes cached input; fresh input excludes cache hits. These three numbers are not additive. Partial metering displays only the measured subset, with coverage disclosed. Unknown never means zero. Native reports and server windows are overlapping scopes, not extra tokens.

The recorded work

The measurement download contains the task identity, outcomes, timings and token receipts used by this page. Reviewed work records provide the associated actions, checks and delivered files. Qualification history contains selected editions, not every development experiment; earlier included failures are retained.

The page runs without telemetry. Source hashes bind the displayed measurements to their saved records.

Source summary SHA-256
e60d97dad8046fa872791b89b9a12c2d1f11d37c62b35c72d84d998cffcc0619

Put your model to work.

Your model.
A better factory.

Get BANTAM FACTORY ↗

Recorded results, ready to explore.