Decision model

GLiDE 0.4B

1 question per request

Accuracy, all 2,198 questions
22.2%20.6% to 24.0%, 95% interval; 489 right
Answers kept at 99% confidence
No answers at this thresholdnone of the 2,198 questions at 99% or more
Cost, $ per 1,000 questionsper 1M questions
No list priceBasis and price under Cost
Latency
0.016 smedian request, 1 question per request

Back to the leaderboard.

How we called it

Kind
Decision model
Access
GLiDE server (Fastino's release code), self-hosted
Paid as
Our own hardware
Model ID
glide-0.4b
Variant
local
Parameters
400M dense, as the maker states it (counted from the released files: 0.40B).encoder (no MLM head) + 255-way readout; source: counted from the checkpoint's safetensors, fastino/GLiDE-0.4B@6bda7ab, huggingface.co/fastino/GLiDE-0.4B; read 10 November 2026
Hardware
1 × DGX Spark, GB10 (128 GB unified memory); BF16
Note
Fastino's released GLiDE 0.4B (fastino/GLiDE-0.4B at 6bda7ab), an encoder (ModernBERT architecture, from Ettin-encoder-400m) that reads at most 7,999 tokens, run on our hardware through the server it ships.
Confidence from
Option score
Run
931, 10 November 2026

Settings

Reasoning
off
Request format
1 question per request
Requests at once
1
Temperature
not recorded
Run record: harness seed, harness commit and prompt version
Harness seed
0
Harness commit
3b14502cefed
Prompt version
254597b8be16

Results

QuestionsAccuracy95% intervalRight
All questions22.2%20.6% to 24.0%489 of 2,198
Reading24.5%21.4% to 28.0%161 of 656
Grammar21.3%19.3% to 23.4%328 of 1,542

Comparison group: ΔΘΛΞ. Compare with another setup.

Accuracy by topic
TopicPartAccuracyRight
Word and phrase meaningsReading22.1%29 of 131
Sentence meaningsReading28.6%44 of 154
How a writer explains and develops ideasReading15.7%11 of 70
Topic and main point of a paragraphReading31.6%31 of 98
How a paragraph fits togetherReading23.6%25 of 106
Details that support a paragraph’s main pointReading21.6%21 of 97
Word types, such as nouns and verbsGrammar25.0%35 of 140
Groups of words built around a nounGrammar13.8%9 of 65
Verbs and their formsGrammar22.0%20 of 91
Verb forms used like nouns or modifiersGrammar19.4%14 of 72
Word endings and what they doGrammar22.8%18 of 79
How words are builtGrammar20.3%14 of 69
The roles of words and phrases in a sentenceGrammar22.1%32 of 145
How verbs show who acts and who receives the actionGrammar20.3%15 of 74
Types of sentencesGrammar17.6%15 of 85
Questions combining several grammar topicsGrammar16.8%37 of 220
Sound patterns and changesGrammar15.6%12 of 77
SpellingGrammar23.3%31 of 133
PunctuationGrammar26.5%31 of 117
Problems in sentence wordingGrammar25.7%45 of 175

English

The Text tab's 547 kept reading questions translated into English, asked with this setup's Text tab settings. Every setup: English reading; against Turkish: English and Turkish.

Right answers
325 of 547, first English run (933)
Accuracy
59.4%
Non-hallucination with “I don't know”
53.1%: 153 “I don't know” of 288 not right (run 935)
Answered with “I don't know”
72.0%: 394 of 547
Error rate with “I don't know”
34.3%: 135 of 394 answers
Cost, $ per 1,000 questionsper 1M questions, first English run
No list price
Latency, median request, first English run
0.014 s; 1 request at once
English run
933
Machine, English runs
DGX Spark
Date
10 November 2026
Request format
1 question per request
Reasoning
off
Prompt version
47f5776e71db
Translations, SHA-256
bf0d47cd385a93050c65906977e199ef7e45d5848dbbdef2bc8d124f4d5abf68

Confidence

Answers kept at 99% or more
No answers at this threshold; 2,198 skipped, the model less than 99% sure
Calibration gap
8.0 points
AUROC, all questions
0.52 0.49 to 0.54
AUROC, reading
0.52 0.47 to 0.57
AUROC, grammar
0.51 0.47 to 0.54
Brier score
0.184
Accuracy of its surest answers
most confident fifth: 23.4% right; most confident half: 22.7% right; most confident four fifths: 23.0% right
Typical score of a right answer
28.2%
Typical score of a wrong answer
28.1%

Repeats

We did not repeat this setup.

Latency

Latency, median request
0.016 s (this run)
Latency, 90th percentile
0.021 s
Requests at once
1
Throughput
3608.5 questions per minute of request time
Request time
0.6 minutes, over the requests whose answers were kept

Latency: the time from sending one request to receiving its answer, measured from our computer, network and any retries within the request included. Request time: the time at least one request was in flight, over the requests whose answers were kept. It is not the run's elapsed time: it leaves out requests that failed and any pause before a run was resumed.

Cost

Cost per 1,000 questionsper 1M questions
No list price
Basis
No dollar estimate; see hardware and time
Input tokens per question
526
Output tokens per question
0

Going first in front of an LLM

Replay of recorded answers, cutoff chosen on half the questions, averaged over the splits. Savings are estimates from API prices; the Method page explains how.

LLM behind itShare takenAccuracy changeSaving
Gemini 3.8 Flash high · chat0.0%−0.02 points0%†
GPT-6 Astra high · chat0.0%−0.02 points0%†
GPT-6.1 Sol high · chat0.0%−0.02 points0%†
GPT-6 Astra low · chat0.0%−0.02 points0%†
GPT-6 Astra low · chat · 1 per request0.0%−0.02 points0%†
Gemini 3.8 Flash low · chat · 1 per request0.0%−0.02 points0%†
GPT-6.1 Sol low · chat · 1 per request0.0%−0.02 points0%†
GPT-6.1 Sol low · chat0.0%−0.02 points0%†
Claude Opus 5.5 high · chat0.0%−0.02 points0%†
GPT-6.1 Terra high · chat0.0%−0.02 points0%†
Claude Opus 5.5 low · chat · 1 per request0.0%−0.02 points0%†
Gemini 3.8 Flash low · chat0.0%−0.02 points0%†
GPT-6.1 Terra low · chat0.0%−0.02 points0%†
Claude Opus 5.5 low · chat0.0%−0.02 points0%†
Claude Sonnet 5.5 high · chat · 1 per request0.1%+0.00 points0%†
Claude Sonnet 5.5 high · chat0.0%−0.03 points0%†

† Leaves out the first model’s hardware cost, which we have not estimated.