PreciseCoder-27B has 26.896 billion total parameters. The selected Q4_K_M GGUF file is 16.547 GB.
Data snapshot: . 2 published tests in this catalog.
Model file and details
Provider
PreciseCoder
Total parameters
26,895,998,464
Exact file size
16,547,399,808 bytes
Quantization / precision
Q4_K_M
Weight repository
mradermacher/PreciseCoder-27B-GGUF
Original model repository
PreciseCoder/PreciseCoder-27B
Weight revision
b74b674cf3355e60836a94c6ef09dbc172a07112
Model revision
4040b8a6b8ea57ce64e94ff65829cff5dc5d9f0c
Runtime notes
Confirm model architecture and quantization support in your runner.
Recorded release date
2026-10-07
Date evidence
PreciseCoder published its Apache-2.0 Qwen3.6-27B debugging specialist on 7 October 2026 and identified checkpoint 328 after two GDPO epochs as the best validation checkpoint. The selected third-party Q4_K_M GGUF is a complete 16,547,399,808-byte language-weight file. Publisher evaluation results apply to the source checkpoint, not this quantization; no quantization-fidelity, inference-speed or minimum-memory result is published. The file size excludes runtime, KV cache, context and output headroom, and optional vision projectors are separate.
These scores were compiled from the linked sources. These published scores were not independently tested here. Different prompts, precision, thinking budgets and tools can change results.
PreciseCoder · PDB test unit pass
79.3 / 100
Publisher-reported unit-test pass rate on PDB test buggy inputs. This is a specialist program-debugging evaluation, not a general coding leaderboard.
Benchmark efficiency: 4.792 published score points per GB of selected weights. This ratio is not an inference-speed measure; check the test settings to see whether the score covers the selected quantization.
Test settings: Publisher evaluation of the source checkpoint on PDB test buggy inputs with a 4,000-token thinking budget and 2,000-token answer budget. The card reports 0.793 unit-test pass, 0.792 edit precision, 0.867 edit recall, 0.810 edit F1 and 2.6 edited lines on average.
Publisher-run source-checkpoint result, not an independent validation or a measurement of the selected Q4_K_M GGUF. The card does not identify this as a general coding benchmark. Training used PDB's train split with group-relative rollouts, so the held-out split and exact harness matter. Apache-2.0 license.
Publisher-reported unit-test pass rate on 228 PDB-Wild SWE-smith repository-level bugs. This is a specialist program-debugging evaluation.
Benchmark efficiency: 3.155 published score points per GB of selected weights. This ratio is not an inference-speed measure; check the test settings to see whether the score covers the selected quantization.
Test settings: Publisher evaluation of the source checkpoint on PDB-Wild, a 228-example set of SWE-smith repository-level bugs. The card reports 0.522 unit-test pass, 0.551 edit precision, 0.629 edit recall, 0.566 edit F1 and 9.2 edited lines on average.
Publisher-run source-checkpoint result, not an independent validation or a measurement of the selected Q4_K_M GGUF. Repository selection, generated bugs, test coverage, prompt format, reasoning budget and runtime can affect the result. The publisher separately reports a 45.8% no-op rate and 11.1% damage rate on 1,512 clean gt_holdout inputs, so success on buggy inputs should not be read as proof that edits are always safe.
Context and version evidence: The repository and file name identify the selected model. Revision hashes are shown above when recorded. Tested context length, full hardware configuration and run date are not consistently recorded in this snapshot. Only settings explicitly stated in each benchmark record are known.
The 16.547 GB figure covers model weights in this file. It excludes the context cache, runtime, operating system and any separate multimodal files. It is not a measured RAM requirement or a token-generation speed.
These complete configurations pass the initial weight-size check for the selected 16.55 GB file. Allow extra memory for context and software. Check the model’s runtime notes and your inference software before buying.
Listing evidence: 2026-09-25. Amazon.com base price visible in an owner-supplied screenshot reviewed 25 Sep 2026; capture time not supplied. Dated reference, not a live quote. Excludes card signup offers, coupons, tax and shipping.
Listing evidence: 2026-10-03. Apple US launch price announced 22 September 2026: USD 1,699. Dated manufacturer reference, not an Amazon offer; tax and shipping may differ.
Listing evidence: 2026-09-24. GMKtec, EVO-X2 / 64GB / 1TB, recorded 22 Sep 2026; exact ASIN capacity identified 24 Sep. This is a dated reference, not a current Amazon price; tax, shipping and bundle contents may differ.
Memory matches are estimates, not tested compatibility or speed. As an Amazon Associate I earn from qualifying purchases.
Confirm model architecture and quantization support in your runner. The selected weights use 16.547 GB; this excludes context cache, runtime buffers and other programs.
These are the first listed tiers that pass our reserve policy, not measured minimum requirements. System RAM and VRAM remain separate. Use the memory tool to change the reserve for your context and runner.