FreshContextA context lab · Fife
Drawing
02 · Benchmarks
Sheet
2 of 3
Rev
A · 2026-10-01

Model← Models

GLM-4.7-Flash

31B mixture of experts, ~3B active, Q8_0, 30.0 GB.

Parameters
31B / ~3B active
Quant
Q8_0
Size
30.0 GB
Family
glm
RecordStatus · runtime · seat

glm-4.7-flash-q8

Measured, not seatedMoEresidentllama.cppmlx

Model card ↗

Outcome as of 20 September 2026, from the seat registry. The write-up below was made when the model was tested.

Why it was tested

Tested as a fast MoE with a 202k context window, and as the first model we quantised ourselves for MLX.

What came of it

Weak at abstention: 48.6% nullschema means it invents values on half the cases where a field is absent, and 4-bit costs a further 19 points over 8-bit. But its context behaviour is strong and was initially misjudged here: it holds 131k in 36.42 GB total at 37.2 tok/s. A first assessment predicted 88 GB from its head counts and was wrong by 7x. Good for long-context work that does not depend on declining to answer.

Recorded runs3 scored · 95% intervals
SuiteScore95% interval PassedRuntime · profileRecorded
grounding76.0%67.8–82.6Passed 95/125llama.cppextraction2026-08-26
nullschema48.6%37.4–59.9Passed 35/72llama.cppextraction2026-08-26
screening95.0%86.3–98.3Passed 57/60llama.cppextraction2026-08-26

Runs marked unenforced were measured through mlx_lm on a suite that requires constrained decoding. mlx_lm accepts the schema and ignores it, so those numbers do not measure what the suite is designed to measure. They are kept and labelled rather than deleted.

Every run log entry for this model (5)
FinishedSuiteRuntime · profileScoreCoverageGrader · dataset · config
nullschemanullschema/1llama.cppextraction48.6%Coverage 72/72g d7448e0f5fc3f9afd 3b91280bb40372efc 3d836e2b73f9cd77
screeningscreening/1llama.cppextraction95.0%Coverage 60/60g 0661b49a10be2d2ad 64c97ca2269c29afc b7d076058290ae5f
groundinggrounding/1llama.cppextraction76.0%Coverage 125/125g 30b933dcc6a7e2d4d dcb396faf8b83b0ac f7116f1eb6a7f001
nullschemanullschema/1llama.cppextractionn/aCoverage 0% coverageg d7448e0f5fc3f9afd 3b91280bb40372efc 3d836e2b73f9cd77
screeningscreening/1llama.cppextractionn/aCoverage 0% coverageg 0661b49a10be2d2ad 64c97ca2269c29afc b7d076058290ae5f