Same model, different agent: 17 points apart on 21 MCP servers
Comments on “Same model, different agent: 17 points apart on 21 MCP servers”
Claim-checked. Scored by all three judges, and its checkable claims were compared with the sources it cites. How far stories are checked
SineFrame, maker of the open-source M3 MCP testing tool, ran Claude Haiku 5.5 through two coding agents against 21 MCP servers. OpenCode answered 82.0% of tasks from the server, Claude Code 65.4%. The gap sits on the ten servers that overlap built-in tools such as web search, where agents often answer without calling the server. On the other eleven, agents scored within about four points. For server builders, results depend on the agent used for testing. The figures are SineFrame's own and unverified; Codex also ran a different model.
Why it ranked
No. 19 of 43 with a score of 0.63. The panel rated it highest on new information (0.89) and substantiated (0.87), lowest on significance (0.52). Placed No. 19, No. 15 by score alone: the eighth AI & machine learning story of the day (−0.080). The placement terms move stories; they are not part of the score.
- AI judgeJudge 2claude-haiku-5-5
- Original benchmark comparing agents' use of MCP servers, with solid numbers and a transparent method. Relevant to builders of AI tooling, but narrow in audience and partly promotional for the authors' own tool.
- AI judgeJudge 3gpt-5.6-luna
- A well-supported comparative measurement of agent behavior across 21 MCP servers, with reproducible methodology, public leaderboard, transcripts, and actionable implications for MCP developers. It is highly relevant, though scope is specialized and one comparison confounds agent with model.
Each judge wrote its line after answering the rubric, with the site and the author still hidden. Every answer behind this is in the table under Score.
Score
0.63total under rubric v2.3
- Significance0.52
- Reader interest0.77
- Audience breadth0.53
- Practical relevance0.77
- Substantiated0.87
- Technical depth0.74
- New information0.89
- Counts against: Promotional0.18
- Counts against: Overclaim0.20
Each judge’s answer per question (0–1), their mean and spread, and what the question adds to the score (weight × mean).
| Question | Judge 1 | Judge 2 | Judge 3 | Mean | Spread | Weight | Contribution (weight times mean) |
|---|---|---|---|---|---|---|---|
| SignificanceHow much does this change what people in tech know or can do? | 0.59 | 0.30 | 0.68 | 0.52 | 0.38 | +0.24 | +0.126 |
| Reader interestWould a well-informed senior engineer or tech founder want to read this today, and likely pass it on? | 0.74 | 0.70 | 0.86 | 0.77 | 0.16 | +0.26 | +0.199 |
| Audience breadthHow many people who build technology does it concern? | 0.55 | 0.45 | 0.58 | 0.53 | 0.13 | +0.13 | +0.068 |
| Practical relevanceHow useful is it to someone who builds software or tech products? | 0.82 | 0.70 | 0.78 | 0.77 | 0.13 | +0.22 | +0.169 |
| SubstantiatedHow well is it backed? Consider whether it is the primary source, and whether its main claims come with data, code, a demo, a reproducible method or citations. | 0.97 | 0.72 | 0.91 | 0.87 | 0.25 | +0.05 | +0.043 |
| Technical depthHow technically substantive is it? | 0.88 | 0.60 | 0.74 | 0.74 | 0.28 | +0.05 | +0.037 |
| New informationDoes it contain information absent from the stories covered in the last 7 days (list provided)? | 0.84 | 0.85 | 0.97 | 0.89 | 0.13 | +0.05 | +0.044 |
| PromotionalIs it mainly marketing with little substance (a press release, ad or sponsored post with no real information)? A launch post that carries real news is not promotional. | 0.08 | 0.35 | 0.12 | 0.18 | 0.27 | −0.19 | −0.035 |
| OverclaimDo the headline or conclusions claim more than the body supports? | 0.21 | 0.30 | 0.08 | 0.20 | 0.22 | −0.11 | −0.022 |
| Sum of contributions | +0.631 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.631 | ||||||
Spread is the highest minus the lowest judge answer. marks a spread above 0.40: the judges disagree, and the story is marked contested.
The badge for this ranking, for the author's page or anyone's: one line of HTML, and it links back here.
<a href="https://rubric.hrih.si/editions/2026-11-02/19/"><img src="https://rubric.hrih.si/badge/2026-11-02/19.svg" alt="Ranked #19 on Rubric, 2 Nov 2026" width="238" height="28"></a>Claim check
Claim check: 2 no source given · 6 couldn’t check. Each claim is compared with the source the article itself cites; a flag is published only when both checkers agree.
Couldn’t checkBenchmark result
Through OpenCode, Claude Haiku 5.5 answered 82.0% of tasks correctly from the server, versus 65.4% through Claude Code.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “Through OpenCode it answered 82.0% of the tasks correctly from the server. Through Claude Code it answered 65.4%”
The source says No passage could be compared.
Source cited by the article: m3.sineframe.com/leaderboard
Couldn’t checkNumber
On the 11 servers with no built-in overlap, the three agents land within 4.1 points of each other.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “On the first eleven, the three agents land within 4.1 points of each other, even though Codex runs a different model”
The source says No passage could be compared.
Source cited by the article: m3.sineframe.com/leaderboard
Couldn’t checkBenchmark result
On the ten servers overlapping built-in tools, Claude Code trails OpenCode by 32 points and Codex trails it by 49.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “On the other ten, Claude Code trails OpenCode by 32 points and Codex trails it by 49”
The source says No passage could be compared.
Source cited by the article: m3.sineframe.com/leaderboard
Couldn’t checkNumber
On the ten overlapping servers, Claude Code answered 378 trials through its own tools, 342 of which contained the expected value, while OpenCode answered 195 that way.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “Claude Code answered 378 trials through its own tools, and 342 of those answers contained the expected value. OpenCode answered 195 that way”
The source says No passage could be compared.
Source cited by the article: m3.sineframe.com/leaderboard
Couldn’t checkNumber
Codex skipped the server on 44.5% of trials where it should have called it, versus 36.5% for Claude Code and 16.1% for OpenCode.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “Codex skipped the server on 44.5% of the trials where it should have called it, against 36.5% for Claude Code and 16.1% for OpenCode”
The source says No passage could be compared.
Source cited by the article: m3.sineframe.com/leaderboard
No source givenNumber
OpenCode never called a built-in web search tool in 1,239 trials, and it scored 96% on Exa while the other two agents scored 11%.
The article gives no link or citation for it.
The article says “In 1,239 trials it never called a built-in web search tool, and the only web built-in it used was webfetch”
The source says No source to compare with.
No source givenCausal conclusion
Most of Claude Code's 17-point deficit is explained by right answers that came from somewhere other than the server.
The article gives no link or citation for it.
The article says “Most of Claude Code's 17-point deficit is right answers that came from somewhere other than the server”
The source says No source to compare with.
Couldn’t checkNumber
The benchmark comprised 3,717 scored trials, run between 29 October and 1 November 2026, scored under spec 1.6.0.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “3,717 scored trials in all. Scoring follows spec 1.6.0”
The source says No passage could be compared.
Source cited by the article: m3.sineframe.com/leaderboard
Also covered by
No other coverage was clustered with this story.
Elsewhere
Elsewhere: HN 2 pts, not on front page
Recorded for context only, checked 2 Nov 2026, 05:46 UTC. It played no part in the ranking.