Ling 3.0 Flash vs MiniMax M2.7
Change modelsLing 3.0 FlashInclusionAI
MiniMax M2.7MiniMax
IndexAll categories, out of 100.
50.9 ±4.7
51.1 ±4.5 best
GapIndex points below the top.
−0.2 · no clear gap
Top
CoverageIndex weight that has results.
85%
85%
SpeedOutput tokens per second.
73/s best
35/s
EloLMArena rating and rank.
—
1405
ReleasedMonth of release.
2026-07
2026-03
Input / 1MUS dollars per 1M input tokens.
$0.021 best
$0.21
Output / 1MUS dollars per 1M output tokens.
$0.063 best
$0.84
1K chats2K tokens in, 500 out, each.
$0.074 best
$0.84
1K documents20K tokens in, 1K out, each.
$0.483 best
$5.04
1K agent tasks200K tokens in, 20K out, each.
$5.46 best
$58.80
ContextMaximum tokens in one request.
262K best
197K
AgenticMulti-step tasks with tools.
53.6 best
52.0
CodingCode writing and repair.
49.9
52.3 best
ReasoningLogic problems and puzzles.
48.7
49.5 best
MultimodalTasks with images and text.
—
—
KnowledgeFacts and expert knowledge.
46.0
50.1 best
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
58.6
59.7 best
MathMath problems.
54.3 best
47.6
Benchmarks
27 sharedOnly Ling 3.0 Flash: 16Only MiniMax M2.7: 26
Ling 3.0 Flash
MiniMax M2.7