ModelsCompareBest forBenchmarksStatusPricingAPI
Compare / head-to-head

GLM-4.7vsLlama 3.3 70B Instruct

GLM-4.7 leads 3 of 3 shared benchmarks. GLM-4.7 has the larger context window (200K tokens).

Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.

Shared benchmarks
3 – 0
GLM-4.7 leads
Cheaper per token
list price, input + output
Larger context
GLM-4.7
200K tokens
Provider uptime (30d)
— · —
Z.ai · Meta
GLM-4.7
Z.ai · released 2025-12-22
text
Context
200K
Max out
131K
Input /1M
$0.6
Output /1M
$2.2
Cached /1M
$0.11
Scores
4 · 4 core
Llama 3.3 70B Instruct
Meta · released 2024-12-06
text
Context
131K
Max out
Input /1M
Output /1M
Cached /1M
Scores
5 · 4 core
Quality

Benchmark matrix

BenchmarkGLM-4.7Llama 3.3 70B InstructΔEdge
Reported by both · 3
GPQA Diamond83.3% epoch50.5% +32.8 ptGLM-4.7
LMArena Elo1435.9 1274 +161.9GLM-4.7
OTIS Mock AIME 2024-2025 (Epoch AI run)83.3% epoch5.1% epoch+78.2 ptGLM-4.7
Only GLM-4.7 reports · 1
SimpleQA Verified32.2% epochnot reported
Only Llama 3.3 70B Instruct reports · 1
MMLU-Pronot reported68.9%
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is GLM-4.7 minus Llama 3.3 70B Instruct in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing

Side by side

SpecGLM-4.7Llama 3.3 70B InstructEdge
Context window200K tokens131K tokensGLM-4.7
Max output131K tokens
Input price / 1M$0.6
Output price / 1M$2.2
Cached input / 1M$0.11
Input + output / 1M
Lower is cheaper. List prices; batch, tool and regional fees excluded.
$2.8
ModalitiestexttextTie
Released2025-12-222024-12-06
Cited benchmark scores45
Reliability

Provider status

All providers
More matchups

Llama 3.3 70B Instruct vs …

Built by Respan
Which one wins on your data?

Public benchmarks are a starting point. Run GLM-4.7 and Llama 3.3 70B Instruct on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.