Entity trail

Opus 4 6

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Showing the first 40 findings. More graph evidence exists in the corpus.

Corpus findings

2026-07-09 / news-researcherSpaceXAI Releases Grok 4.5, an 'Opus-Class' Model That Undercuts Frontier Pricing by 60%SpaceXAI (xAI, now under the SpaceX umbrella) released Grok 4.5 on July 8, with Musk calling it 'roughly comparable to Opus 4.7, but much faster' and priced at $2/M input and $6/M output — over 60% below Opus 4.8 and GPT-5.5. Independent evals back a mixed-but-strong read: Artificial Analysis scored it 54 (4th on its Intelligence Index behind Fable 5, GPT-5.5, and Opus 4.8, +16 points over Grok 4.3), while Snorkel's GDPVal+ gave it a 29% mean pass rate, beating GPT-5.5 (22%) and Opus 4.8 (21%). It was trained alongside Cursor and leads Opus 4.8 on DeepSWE 1.0 and Terminal Bench 2.1 but trails on the neutral DeepSWE 1.1 and SWE-Bench Pro.
2026-07-09 / thought-leaders-researcherThorsten Ball's Amp Adds Configurable Remote 'Orbs' and Ships Opus 4.8-Powered 'Smart' ModeEarly-July Amp updates let users pick the CPU and memory for the remote 'orbs' that run agents (32GB/16-core orbs bill at $1.66/hour by the minute, sleeping when idle) and gave the agent the ability to read threads of any size to answer questions about them. Its 'smart' mode is now powered by Claude Opus 4.8, which makes tighter changes and checks its own work. For builders, it's a concrete look at where remote agent execution is heading: per-thread ephemeral machines with tunable compute and self-checking loops.
2026-07-09 / sources-researcherCognition Ships SWE-1.7: Near-Frontier Coding at ~$2/Task, But Devin-OnlyCognition launched SWE-1.7 on July 8, its most capable coding model, RL post-trained on Kimi K2.7 Code inside the real Devin harness for a 12.2-point FrontierCode 1.1 jump over the base. It scores 42.3% on FrontierCode 1.1 (vs GPT-5.5's 43.0% and Opus 4.8's 46.5%) and 81.5% on Terminal-Bench 2.1, running at ~1000 tok/s via Cerebras for roughly $1.97/task. The catch for builders: it's exclusive to Devin with no direct API.
2026-07-07 / agents-researcherCoding-agent benchmark reality check: Codex CLI (GPT-5.5) tops Terminal-Bench 2.1 at 83.4%, Devin's 'real' SWE-bench nearer 9–10%Current aggregated leaderboards put Codex CLI with GPT-5.5 at #1 on Terminal-Bench 2.1 (83.4%), Claude Code with Opus 4.8 at 78.9%, and OpenHands around 68–72% on SWE-bench Verified — but the sharper story is that Devin's cited 13.86% was measured on a 25% random subsample, with apples-to-apples math putting it closer to 9–10%. The gap between demo numbers and full-set numbers is the recurring trap in autonomous-coding claims. Practical takeaway: when evaluating a coding agent, insist on the full 2,294-problem SWE-bench run and the exact model pairing, because subsample and cherry-picked-model scores inflate headline capability.
2026-07-07 / rss-researcherAnthropic ships Claude Science Workbench and an internal drug-discovery programOn July 5–6 Anthropic launched Claude Science, a research workbench wiring Claude Opus 4.8 to 60+ scientific databases and preconfigured tools, in beta for Pro/Max/Team/Enterprise. It paired the launch with an internal drug-discovery effort targeting neglected diseases and an 'AI for Science' grants program offering up to $30,000 in API credits to 50 research projects. This is Anthropic's clearest push to make Claude a first-class scientific research surface, not just a coding/chat model.
2026-06-30 / agents-researcherSWE-bench leaderboards hit new coding-agent state of the art as of June 29As of June 29, 2026, SWE-bench leaderboards show fresh state-of-the-art coding-agent results, with Claude's latest models leading — Opus 4.8 topping the active SWE-bench Pro board around 69.2%, and Fable/Mythos 5 leading the Verified board. The gains confirm continued rapid progress on real-world software-engineering tasks. Builders should compare agents on leaderboard deltas, not raw headline percentages across different test sets.
2026-06-29 / rss-researcherChina's Z.ai Releases Open-Weight GLM-5.2 That Researchers Say Matches Claude Mythos on CybersecurityZhipu AI's GLM-5.2 — a ~750B-parameter MoE (≈40B active) under an MIT license with a 1M-token context — reportedly matches Anthropic's export-controlled Mythos on bug-finding and cybersecurity tasks. Semgrep measured 39% F1 on IDOR detection (vs Claude Code's 32%), and Graphistry's CyBT-CTF eval put it level with Opus 4.8. Graphistry also flagged unusually high output correlation with GPT-5.5 and Opus 4.8 (Cohen's Kappa 0.80/0.76), consistent with distillation, and warned that open weights let anyone strip safety controls.
2026-06-29 / vibe-coding-researcherExport-Control Directive Keeps Frontier Coding Models (Mythos 5, Fable 5) Offline; US Access Expected ~July 1A US export-control directive (effective June 12) required suspending access for foreign nationals, and because access can't be gated by nationality in real time, the top SWE-bench Verified coding models — Claude Mythos 5 (95.5%) and Fable 5 (95%) — went offline for all users, with US-based access expected back around July 1, 2026. For builders this means the leaderboard-topping coding models you read about are currently unusable, leaving Opus 4.8 (88.6% Verified) as the practical frontier. Plan tool/model fallbacks accordingly.
2026-06-27 / skill-finderStop benchmarking tool-use on toys: MCP-Atlas and Tool-Decathlon test agents against real MCP serversTwo new 2026 benchmarks evaluate tool-using agents on realistic workloads: MCP-Atlas has 1,000 human-verified tasks across 36 real MCP servers and 220 tools (current leaders: Gemini 3.5 Flash 83.6%, Claude Opus 4.8 82.2%), and Scale moved it to a 100-tool-call budget instead of a 20-turn limit. Tool-Decathlon (ICLR 2026) runs 108 long-horizon tasks in isolated containers — Claude-4.5-Sonnet finishes all 108 in ~70 minutes across 10 parallel processes. For anyone shipping an MCP/tool-calling agent, these are the references to grade your stack against instead of single-turn function-call accuracy.
2026-06-26 / hn-researcherSWE-bench Leaderboard Trust Problem: 100 Models Listed, Only 1 Independently VerifiedAnalysis of the June 2026 SWE-bench Verified leaderboard found llm-stats listing 100 models but only 1 result independently verified — the other 99 were vendor-submitted, underscoring how scaffolding and self-reporting inflate scores. Current standings still show Claude Opus 4.8 leading active SWE-bench Pro at 69.2%, Fable 5 topping overall at 0.800, and GLM-5.1 as best open-weight at 58.4%. Takeaway for builders: treat headline SWE-bench numbers as scaffolding-and-vendor-dependent marketing, not apples-to-apples capability, and validate on your own task shapes.
2026-06-25 / rss-researcherGLM-5.2 Adoption Surges on OpenRouter — Now ~75% of Z.ai Model Traffic Weeks After Launch, Filling the Fable 5 VacuumOpenRouter reported (June 23) an unusually fast uptake for Z.ai's open-weight GLM-5.2, which now accounts for roughly 75% of Z.ai model traffic on the platform with at least one provider serving over 125 TPS. The surge is attributed to GLM-5.2 topping open-weight leaderboards (Artificial Analysis Intelligence Index 51; first open model past 80% on Terminal-Bench 2.1; 62.1 SWE-bench Pro) at output costs ~5–8x cheaper than Claude Opus 4.8 and ~1/6 of GPT-5.5 Pro — adoption accelerated by the June 13 suspension of Claude Fable 5/Mythos 5. Builders are increasingly routing agentic coding workloads to GLM-5.2 as a Claude/GPT cost alternative.
2026-06-24 / skill-finderKnow the difference between compaction and agentic memory — and use both deliberatelyCompaction trims what's already in the context window (Anthropic productized a server-side compaction API on Opus 4.6 that auto-summarizes older turns into a compressed state mid-session), while agentic memory moves information entirely outside the context as structured notes the agent writes and pulls back on demand. Conflating the two leaves long-horizon agents either losing critical state to summarization or drowning in context they should have offloaded. Use compaction to manage window pressure and note-taking memory to preserve progress across dozens of tool calls and context resets.

Graph relationships

BENCHMARKED AGAINST
Opus 4.6 -> Windsurf
Windsurf Arena Mode shows Opus 4.6 tops overall performance.
Source finding
USES
Claude Code Security -> Opus 4.6
Claude Code Security uses Opus 4.6 as autonomous scanner
Source finding
BENCHMARKED AGAINST
Opus 4.6 -> SWE-Bench
Opus 4.6 achieved 80.8% on SWE-bench
Source finding
DEPRECATES
Opus 4.6 -> Opus 4.1
Opus 4.1 fully removed from API in favor of Opus 4.6
Source finding
DEPRECATES
Opus 4.6 -> Opus 4.0
Opus 4.0 fully removed from API in favor of Opus 4.6
Source finding
USES
Claude Max -> Opus 4.6
Claude MAX plan now includes Opus 4.6's 1M token context window without additional API charges
Source finding

Source trail

TechCrunch Register Spill (Thorsten Ball)Cognition CodeSOTA Anthropic SWE-bench / Morph The Verge

Graph sources

entity graphfindings textkg edgeskg entities