LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

Cheapest good LLM for knowledge

GLM 5.3 Flash leads
$0.50
models qualify
46

Output $/Mtok · read October 21

Cheapest output, $ per million tokens

  • Mistral Nemo $0.03
  • Gemma 3 4B $0.05
  • S Llama 3 8B Lunaris $0.05
  • Ling 3.0 Flash $0.063
  • Mistral Small 3 $0.08
  • Llama 3.1 8B Instruct $0.08
  • gpt-oss-20b $0.09
  • N Nex-N2.5-Mini $0.10

As of October 21, 2026, Z.ai GLM 5.3 Flash is the cheapest model at or above the median MMLU Pro result, at $0.50 per million output tokens, ahead of MiniMax M2.1 ($1.20) and OpenAI GPT-5.6 Luna ($1.20).

This list narrows the best LLM for knowledge ranking.

Top 20 of 46, lower is better

Embed this data
Cheapest good LLM for knowledge
Rank Model Output $/Mtok Coding evidence MMLU Pro Coding index
1
$0.50
86.1% 71.0
2
MiniMax M2.1 MiniMax
$1.20
87.0% Not benchmarked
3
GPT-5.6 Luna OpenAI
$1.20
86.0% 61.5
4
$1.32
86.2% Not benchmarked
5
GLM 5.3 Z.ai
$1.40
86.8% 73.1
6
$1.50
86.2% 8.1
7
$1.80
87.2% Not benchmarked
8
$1.9105
87.3% 35.3
9
$1.95
87.7% 37.4
10
$1.98
87.0% 54.6
11
Grok 4.20 xAI
$2.50
86.3% Not benchmarked
12
$3.00
88.6% 75.8
13
Kimi K2.6 Moonshot AI
$3.41
87.6% 44.2
14
$3.75
90.1% 69.9
15
$3.75
89.3% 48.3
16
T Inkling Thinking Machines
$4.05
86.3% 49.7
17
$4.125
90.2% 70.5
18
$4.25
88.7% 48.9
19
$4.25
88.3% 65.1
20
GLM 5.1 Z.ai
$4.40
86.9% 44.0

How this list is built

Generated from the dataset on every refresh. Nobody picks the order by hand.

Models whose MMLU Pro result is at or above 86.0%, the median of every model on that board, ordered by output price, cheapest first. Free listings are left out: they are a tariff, not a price.

Other lists

Rankings beyond coding · Methodology