Claude Opus 4.5 Thinking vs GLM-5.1
Change modelsClaude Opus 4.5 ThinkingAnthropic
GLM-5.1Z.AI
IndexAll categories, out of 100.
56.3 ±4.2 best
56.0 ±4.1
GapIndex points below the top.
Top
−0.3 · no clear gap
CoverageIndex weight that has results.
95% best
85%
SpeedOutput tokens per second.
—
102/s
EloLMArena rating and rank.
—
1463
ReleasedMonth of release.
2025-11
2026-04
Input / 1MUS dollars per 1M input tokens.
—
$1.40
Output / 1MUS dollars per 1M output tokens.
—
$4.40
1K chats2K tokens in, 500 out, each.
—
$5.00
1K documents20K tokens in, 1K out, each.
—
$32.40
1K agent tasks200K tokens in, 20K out, each.
—
$368.00
ContextMaximum tokens in one request.
200K
205K best
AgenticMulti-step tasks with tools.
59.7 best
59.2
CodingCode writing and repair.
54.4
57.0 best
ReasoningLogic problems and puzzles.
54.2 best
50.9
MultimodalTasks with images and text.
57.3
—
KnowledgeFacts and expert knowledge.
59.0 best
51.7
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
43.2
60.3 best
MathMath problems.
57.2 best
54.9
Benchmarks
21 sharedOnly Claude Opus 4.5 Thinking: 16Only GLM-5.1: 39
Claude Opus 4.5 Thinking
GLM-5.1