Compare / Claude Haiku 5.5 vs MiniMax M3.1 Flash

Claude Haiku 5.5 vs MiniMax M3.1 Flash

Which is cheaper, which is more capable, and which does better at each kind of work, from 89 results on 6 sources that measured both models.

AnthropicClaude Haiku 5.5 Released 7 Oct 2026 · $0.10 in · $0.50 out
MiniMaxMiniMax M3.1 Flash Released 19 Oct 2026 · $0.15 in · $0.60 out
Claude Haiku 5.5 wins 4112 too close to callMiniMax M3.1 Flash wins 36

Of the 89 results both models have, the better value on each. Results whose reported ranges overlap are too close to call.

Results as of 19 November 2026 (catalogue release 2026-11-19-3608133e8331). See also Anthropic vs MiniMax, every model from both labs.

Claude Haiku 5.5 or MiniMax M3.1 Flash? The short answer

They score and cost about the same; Claude Haiku 5.5 writes faster. Pick on the results below that match your work.

  • Spring Prompt overall: Claude Haiku 5.5 38, MiniMax M3.1 Flash 40 (out of 100; how the score works).
  • They cost about the same to run: $0.018 and $0.020 per survey plan and analysis on SurveyBench.
  • Claude Haiku 5.5 scores higher on the Artificial Analysis Intelligence Index: 43.4 against 42.4.
  • Claude Haiku 5.5 writes faster: 240 tokens a second against 159.
  • MiniMax M3.1 Flash is newer: released 19 Oct 2026, against 7 Oct 2026 for Claude Haiku 5.5.
  • On our own benchmarks, on the results both models have: Claude Haiku 5.5 does better for user surveys; MiniMax M3.1 Flash for product listings and marketing planning.
  • On third-party benchmarks, on the results both models have: Claude Haiku 5.5 does better for agents and tool use.

Head to head

38
Spring Prompt overall
40
43.4
Intelligence Index
42.4
$0.20
Price per million tokens, blendedlower is better
$0.26
240
Output speed, tokens a second
159
1,000,000
Context window, tokens
1,048,576

Every fact side by side

Claude Haiku 5.5MiniMax M3.1 Flash
Price per million tokens$0.10 in · $0.50 out$0.15 in · $0.60 out
Blended (3 in : 1 out)$0.20$0.26
Cheapest host, blended$0.20 (Amazon Bedrock)$0.20 (GMICloud, FP8)
Cost per survey plan and analysis (SurveyBench)$0.018$0.020
Cached input, per million–$0.015 cached input
Intelligence Index43.442.4
Spring Prompt overall3840
Output speed, tokens a second240159
Context window1,000,000 tokens1,048,576 tokens
Released7 Oct 202619 Oct 2026 (newer)
WeightsClosedOpen
Developer based inthe USChina

Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.

Which is better for what

Each model's rank among the models on sale today, for each kind of work, from our own benchmarks and licensed sources. Further right is better; the better rank is in green.

Product listingsTurning a sparse product feed and photos into listings that can go live 23 of 35 14 of 35
Decks from an analysisTurning a finished analysis into a deck you could present as it is 19 of 35 10 of 35
User surveysPlanning a user survey and reading its results without being misled 9 of 26 26 of 26
Marketing planningPlanning a year of ad spend without overspending 33 of 35 24 of 35
Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls 28 of 159 33 of 159
Reasoning and knowledgeHard questions across science, maths and general knowledge 25 of 187 19 of 187
Speed under pressureGood decisions against a real clock (fast chess) 24 of 38 34 of 38

Their slides, side by side

The same brief for both: Manchester store investment review for Hearthside Coffee, an invented company. First slide of each deck from DeckBench.

First slide of Claude Haiku 5.5's deck for Hearthside Coffee
Claude Haiku 5.5 · deck rating 1,120 · needs work
First slide of MiniMax M3.1 Flash's deck for Hearthside Coffee
MiniMax M3.1 Flash · deck rating 1,266 · needs work

Every slide, side by side →

From our research

Every shared result

Each model's best result on every measure both have, grouped by benchmark, with the setting that produced it. Under each measure, every model that benchmark has measured is a grey tick, best to the right, so you can see where the two sit in the field. The headline measures of each benchmark come first; the rest are under "Show all". Green marks the better value; where a source reports ranges that overlap, neither is marked.

MeasureClaude Haiku 5.5MiniMax M3.1 Flash
BulletBench measured by Spring Prompt · 3 results
Cost per game · Bullet 60sUS dollars, lower is better 23 · 20 of 62 measured $0.0019no reasoning $0.0016low reasoning
Ladder Elo · Bullet 60sladder Elo, higher is better 50 · 58 of 62 measured 310no reasoning 115low reasoning
Games lost on time · Bullet 60s% of games, lower is better 36 · 52 of 62 measured 37.5%no reasoning 75.0%low reasoning
CatalogBench measured by Spring Prompt · 3 results
Cost per productUS dollars, lower is better 4 · 1 of 36 measured $0.0012 $0
Reliably publish-ready% of products, higher is better 26 · 17 of 35 measured 26.8% 50.0%
Reliably publish-ready · sales brief% of products, higher is better 20 · 10 of 35 measured 10.7% 37.5%
DeckBench measured by Spring Prompt · 3 results
Cost per deckUS dollars, lower is better 4 · 1 of 36 measured $0.0091 $0
Deck ratingrating, higher is betterThe rating weighs each win by how strong the other deck was; the win rate counts wins against whichever decks a model happened to meet, so the two can disagree. 19 · 10 of 36 measured 1,120 1,266
Accurate decks% of tasks, higher is better 22 · 6 of 36 measured 33.3% 83.3%
ROASBench measured by Spring Prompt · 3 results
Overall scorescore out of 100, higher is better 34 · 24 of 36 measured 24.1 40.3
Audience scorescore out of 100, higher is better 30 · 22 of 36 measured 48.4 60.8
Business scorescore out of 100, higher is better 32 · 22 of 36 measured 13.5 38.9
SurveyBench measured by Spring Prompt · 3 results
Analysis ratingrating, higher is betterThe rating weighs each win by how strong the other analysis was; the win rate counts wins against whichever analyses a model happened to meet, so the two can disagree. 9 · 27 of 27 measured 1,233 428
Cost per taskUS dollars, lower is better 4 · 5 of 27 measured $0.0185 $0.0198
Analysis scorepoints of 100, higher is better 11 · 27 of 27 measured 88.4 13.6
Artificial Analysis reported by Artificial Analysis · 3 results
Artificial Analysis Intelligence Indexindex score, higher is better 20 · 21 of 249 measured 43.4max reasoning 42.4
Humanity's Last Exam% of questions, higher is better 25 · 15 of 250 measured 44.4%max reasoning 48.0%
Terminal-Bench 4.0% of tasks, higher is better 20 · 23 of 120 measured 32.8%max reasoning 25.8%
Show all 89 results (71 more)
MeasureClaude Haiku 5.5MiniMax M3.1 Flash
BulletBench measured by Spring Prompt · 2 results
Invalid moves · Bullet 60s% of moves, lower is better 1 · 1 of 62 measured 0.0%no reasoning 0.0%low reasoning
Median move time · Bullet 60smilliseconds, lower is better 43 · 56 of 62 measured 1.7 sno reasoning 3.0 slow reasoning
CatalogBench measured by Spring Prompt · 28 results
Channel compliance% of products, higher is better 1 · 1 of 36 measured 100.0% 100.0%
Channel rules broken% of products, lower is better 22 · 1 of 36 measured 0.6% 0.0%
Channel rules broken · sales brief% of products, lower is better 1 · 1 of 36 measured 0.0% 0.0%
Claims to check% of products, lower is better 25 · 1 of 34 measured 0.6% 0.0%
Claims to check · sales brief% of products, lower is better 24 · 1 of 34 measured 1.2% 0.0%
Conflicts caught% of conflicts, higher is better 1 · 1 of 36 measured 100.0% 100.0%
Content quality% of checks, higher is better 29 · 27 of 36 measured 95.6% 96.2%
Content quality · sales brief% of checks, higher is better 24 · 24 of 36 measured 96.3% 96.3%
Decision accuracy% of decisions, higher is better 28 · 24 of 36 measured 95.6% 96.1%
Failed outputs% of products, lower is better 1 · 31 of 36 measured 0.0% 1.2%
Failed outputs · sales brief% of products, lower is better 1 · 31 of 36 measured 0.0% 1.2%
Field accuracy% of missing fields, higher is better 34 · 27 of 36 measured 91.0% 92.7%
Invented values% of filled values, lower is better 11 · 34 of 36 measured 0.8% 3.8%
Missing UK information% of products, lower is better 1 · 1 of 36 measured 0.0% 0.0%
Missing UK information · sales brief% of products, lower is better 1 · 1 of 36 measured 0.0% 0.0%
Not findable% of products, lower is better 34 · 1 of 36 measured 11.9% 0.0%
Not findable · sales brief% of products, lower is better 34 · 1 of 36 measured 10.7% 0.0%
Publish-ready listings% of products, higher is better 26 · 20 of 36 measured 47.0% 63.1%
Publish-ready listings · sales brief% of products, higher is better 24 · 12 of 36 measured 20.2% 57.1%
Time per productseconds, lower is better 16 · 27 of 36 measured 11.0 25.2
Unsupported claimsclaims per product, lower is better 31 · 15 of 36 measured 1.20 0.39
Unsupported claims% of products, lower is better 28 · 12 of 36 measured 32.1% 7.7%
Unsupported claims · sales briefclaims per product, lower is better 25 · 11 of 36 measured 2.46 0.60
Unsupported claims · sales brief% of products, lower is better 25 · 9 of 36 measured 70.2% 15.5%
Wrong attributes% of products, lower is better 26 · 32 of 36 measured 19.6% 22.6%
Wrong attributes · sales brief% of products, lower is better 22 · 30 of 36 measured 18.5% 21.4%
Wrong category or variant% of products, lower is better 1 · 23 of 36 measured 0.0% 2.4%
Wrong category or variant · sales brief% of products, lower is better 1 · 32 of 36 measured 0.0% 4.8%
DeckBench measured by Spring Prompt · 14 results
Caveat dropped% of tasks, lower is better 16 · 16 of 36 measured 16.7% 16.7%
Clean layout% of tasks, higher is better 10 · 31 of 36 measured 66.7% 0.0%
Design quality% of the maximum, higher is better 22 · 19 of 36 measured 61.1% 66.9%
Draft figure quoted% of tasks, lower is better 21 · 1 of 36 measured 33.3% 0.0%
Findings missing% of tasks, lower is better 25 · 1 of 36 measured 16.7% 0.0%
Head-to-head win rate% of comparisons, higher is betterThe rating weighs each win by how strong the other deck was; the win rate counts wins against whichever decks a model happened to meet, so the two can disagree. 15 · 14 of 36 measured 62.1% 63.9%
Layout defects% of tasks, lower is better 10 · 31 of 36 measured 33.3% 100.0%
Misleading metric used% of tasks, lower is better 31 · 1 of 36 measured 16.7% 0.0%
Presentable decks% of tasks, higher is better 9 · 17 of 36 measured 33.3% 0.0%
Recommendation late or wrong% of tasks, lower is better 22 · 1 of 36 measured 33.3% 0.0%
Slides needing work% of tasks, lower is better 5 · 19 of 36 measured 16.7% 50.0%
Time per deckseconds, lower is better 7 · 26 of 36 measured 62.1 217
Unsupported claims% of tasks, lower is better 26 · 1 of 36 measured 33.3% 0.0%
Unsupported numbers% of tasks, lower is better 28 · 1 of 36 measured 16.7% 0.0%
ROASBench measured by Spring Prompt · 6 results
Consistency scorescore out of 100, higher is better 33 · 29 of 36 measured 10.6 17.3
Contribution profitsimulated US dollars, higher is better 32 · 13 of 36 measured $32,146 $693,379
Cost of a runUS dollars, lower is better 6 · 1 of 36 measured $0.0313 $0
Months over budgetmonths of 12, lower is better 1 · 1 of 36 measured 0 0
Planning scorescore out of 100, higher is better 21 · 28 of 36 measured 54.8 54.4
Return on ad spendprofit per $1 spent, higher is better 32 · 19 of 36 measured $0.03 $0.52
SurveyBench measured by Spring Prompt · 17 results
Head-to-head win rate% of comparisons, higher is betterThe rating weighs each win by how strong the other analysis was; the win rate counts wins against whichever analyses a model happened to meet, so the two can disagree. 8 · 27 of 27 measured 59.9% 20.3%
Leading or double-barrelled questions% of tasks, lower is better 12 · 15 of 27 measured 33.3% 44.4%
Mistaken next steps% of tasks, lower is better 13 · 24 of 27 measured 22.2% 77.8%
Numbers right% of answers, higher is better 1 · 1 of 27 measured 100.0% 100.0%
Objectives unanswerable% of tasks, lower is better 1 · 27 of 27 measured 0.0% 44.4%
Precision ignored% of tasks, lower is better 1 · 21 of 27 measured 0.0% 44.4%
Research scorepoints of 100, higher is better 10 · 27 of 27 measured 90.2 28.3
Sound research% of tasks, higher is better 13 · 17 of 27 measured 11.1% 0.0%
Stakeholder's view presumed% of tasks, lower is better 1 · 25 of 27 measured 0.0% 33.3%
Survey plan scorepoints of 100, higher is better 8 · 26 of 27 measured 94.5 62.5
Survey structure% of tasks, lower is better 12 · 25 of 27 measured 22.2% 44.4%
Time per taskseconds, lower is better 13 · 21 of 27 measured 147 352
Traps handled% of traps, higher is better 11 · 22 of 27 measured 94.2% 75.0%
Traps missed% of tasks, lower is better 4 · 19 of 27 measured 22.2% 88.9%
Unsupported findings% of tasks, lower is better 16 · 22 of 27 measured 77.8% 88.9%
Wrong numbers% of tasks, lower is better 1 · 24 of 27 measured 0.0% 88.9%
Wrong recommendation% of tasks, lower is better 5 · 25 of 27 measured 11.1% 88.9%
Artificial Analysis reported by Artificial Analysis · 4 results
Long-context reasoning (AA-LCR)% of questions, higher is better 21 · 47 of 241 measured 82.7%max reasoning 79.7%
Output speedtokens per second, higher is better 10 · 21 of 78 measured 240max reasoning 159
SciCode% of problems, higher is better 28 · 38 of 121 measured 55.0%max reasoning 53.4%
Time to first answer tokenseconds, lower is better 40 · 51 of 78 measured 6.41low reasoning 13.3

Quick answers

Is Claude Haiku 5.5 better than MiniMax M3.1 Flash?

They score and cost about the same; Claude Haiku 5.5 writes faster. Pick on the results below that match your work.

Which is cheaper, Claude Haiku 5.5 or MiniMax M3.1 Flash?

They cost about the same to run: $0.018 and $0.020 per survey plan and analysis on SurveyBench.

Which is faster, Claude Haiku 5.5 or MiniMax M3.1 Flash?

Claude Haiku 5.5 writes about 240 tokens a second on its usual API, against 159 for MiniMax M3.1 Flash (Artificial Analysis).

Which is better for product listings, Claude Haiku 5.5 or MiniMax M3.1 Flash?

MiniMax M3.1 Flash ranks 14 of 35 models on sale for turning a sparse product feed and photos into listings that can go live, against 23 for Claude Haiku 5.5.

Which is better for decks from an analysis, Claude Haiku 5.5 or MiniMax M3.1 Flash?

MiniMax M3.1 Flash ranks 10 of 35 models on sale for turning a finished analysis into a deck you could present as it is, against 19 for Claude Haiku 5.5.

Which is better for user surveys, Claude Haiku 5.5 or MiniMax M3.1 Flash?

Claude Haiku 5.5 ranks 9 of 26 models on sale for planning a user survey and reading its results without being misled, against 26 for MiniMax M3.1 Flash.

Which is better for marketing planning, Claude Haiku 5.5 or MiniMax M3.1 Flash?

MiniMax M3.1 Flash ranks 24 of 35 models on sale for planning a year of ad spend without overspending, against 33 for Claude Haiku 5.5.

Which has the bigger context window, Claude Haiku 5.5 or MiniMax M3.1 Flash?

MiniMax M3.1 Flash: 1,048,576 tokens, against 1,000,000 for Claude Haiku 5.5.

Run Claude Haiku 5.5 and MiniMax M3.1 Flash on your own prompt.

Benchmarks aren't your task. Test both side by side on your prompt, with every answer checked against your rules and the cost of each one.