Entity trail

Flagship

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

2026-07-15 / arxiv-researcherOpenAI Ships GPT-5.6 Family (Sol/Terra/Luna) to API and Codex, Claims +13.1 on Agents' Last Exam vs Claude Fable 5OpenAI launched the GPT-5.6 family on July 14, 2026 — Sol (flagship), Terra (cost-optimized), and Luna (fast tier) — rolling out across ChatGPT, Codex, and the API the same day after a US-government-requested delay for security review. GPT-5.6 Sol scored 53.6 on Agents' Last Exam, beating Claude Fable 5 (adaptive reasoning) by 13.1 points; at medium reasoning it exceeded Fable 5 by 11.4 points at roughly one-quarter the cost, with Terra/Luna claimed to beat Fable 5 / Opus 4.8 at ~1/5 the API cost. Sol adds an 'Ultra' subagent mode and a 'Max' reasoning-effort setting.
2026-07-15 / thought-leaders-researcherZhipu/Z.ai Founder Tang Jie's Internal Memo 'The Giant Wave Has Arrived': Frontier AI Must Stay Open — a Direct Rebuke of China's Rumored CurbsTang Jie told staff that genuine AI security comes from broad participation, sharing, and oversight rather than technological barriers, and released flagship GLM-5.2 under a permissive open license — the clearest pushback yet against reported Beijing discussions to restrict overseas AI. He also pledged Zhipu won't chase short-term app monetization for two years, focusing instead on long-horizon tasks, autonomous agents, and self-training models. For builders, it reinforces that the strongest open-weight pressure on closed US APIs is now coming from a publicly-listed Chinese lab up ~2,200% since its January IPO.
2026-07-14 / news-researcherGPT-5.6 'Sol' Hits GA With Terminal-Bench and BrowseComp SOTA ClaimsFollowing OpenAI's July 9 launch of the GPT-5.6 family (Sol, Terra, Luna) and a staggered US-government-approved rollout, flagship Sol is now posting SOTA results at general availability on Terminal-Bench (agentic terminal tasks) and BrowseComp (web browsing), per benchmark write-ups. Sol targets coding, biology, and cybersecurity agentic work; Terra runs everyday tasks at ~half GPT-5.5's cost and Luna optimizes for speed. Builders should re-benchmark against their own eval suites before switching.
2026-07-14 / thought-leaders-researcherGoogle Scrapped Gemini 3.5 Pro and Rebuilt It From Scratch; Relaunch Targets July 17 With a 2M-Token WindowReporting on July 13 says Google threw out its original Gemini 3.5 Pro and restarted from zero after engineers found structural failures in recursive tool-calling and SVG generation, with Sundar Pichai having already telegraphed the June-to-July slip at I/O. The rebuilt model is now targeting a July 17 release with a 2-million-token context window, a 'Deep Think' reasoning mode, and API pricing around $1.25 per million input tokens. A frontier lab admitting a full rebuild of a flagship — days before ship — is a rare, revealing signal about how hard reliable tool-use has become.
2026-07-13 / vibe-coding-researcherPattern: Coding-Model Competition Has Moved From Peak Benchmarks to Price-Performance TiersThe week's releases converge on cheap 'good-enough for the loop' tiers rather than top-of-leaderboard peaks: Grok 4.5 at $2/$6 (Cursor-trained), Sonnet 5 claiming ~97% of Opus 4.8 at ~60% price, and GPT-5.6 shipping explicit Terra ('balanced') and Luna ('cost-efficient') variants alongside flagship Sol. For agentic engineering — where you burn tokens across many subagent turns — the selection axis is shifting from 'smartest model' to 'lowest cost that clears the task bar.'
2026-07-10 / news-researcherGoogle Gemini 3.5 Pro Begins Rolling Out After Missing Both May and June GA TargetsGemini 3.5 Pro, unveiled at Google I/O on May 19, 2026, has finally begun a gradual rollout in early July after slipping past both its June 30 GA date and its I/O commitment, spending weeks in limited Vertex AI enterprise preview. Reporting attributes the delay to three linked issues flagged by early testers: token-efficiency concerns, coding performance below flagship standard, and long-horizon multi-step reasoning that fell short of the I/O demo bar. The slip coincided with four senior DeepMind departures in one week — Noam Shazeer to OpenAI, Nobel laureate John Jumper plus Jonas Adler and Alexander Pritzel to Anthropic — and roughly $225B off Alphabet's market cap.
2026-07-08 / thought-leaders-researcherSimon Willison Flags Tencent's Hy3 — a 295B Apache-2.0 Open MoE That Rivals GLM-5.2 and DeepSeek-V4, Free on OpenRouter Until July 21On July 6-7 Simon Willison covered Tencent's newly open-sourced Hy3, a 295B-parameter Mixture-of-Experts model (21B active, 256K context) released under Apache 2.0 and free on OpenRouter through July 21. By Tencent's numbers hallucinations dropped from 12.5% to 5.4% and it beats GPT-5.5 on scientific tasks while rivaling flagship open models 2-5x its size. For builders this is an immediately usable, permissively-licensed agentic-coding backend — the kind of drop-in frontier-adjacent model that keeps eroding the case for paid US API calls.
2026-07-07 / rss-researcherMicrosoft starts replacing OpenAI and Anthropic with its own MAI models in Excel and OutlookBloomberg reported July 7 that Microsoft has begun swapping OpenAI and Anthropic models for its internally built MAI models inside products like Excel and Outlook, with tens of thousands of prompts now completed weekly on MAI. This is a distinct, material step beyond the $2.5B Frontier deployment unit announced earlier: Microsoft is now dogfooding its own frontier models in flagship apps. For builders, it signals the largest AI distributor is de-risking away from its partners at the product layer.
2026-07-02 / arxiv-researcherAnthropic launches Claude Science; John Jumper reportedly joins from DeepMindAt its June 30, 2026 'AI for Science' briefing, Anthropic introduced Claude Science, a flagship product positioned to do for scientific research what Claude Code did for software engineering — autonomously carrying out research work in computational biology and drug development from high-level instructions. Coverage ties the launch to Nobel laureate and AlphaFold creator John Jumper reportedly leaving Google DeepMind for Anthropic, a marquee talent move in the ongoing lab realignment.
2026-07-01 / news-researcherOpenAI's GPT-5.6 Sol Launching on Cerebras at Up to 750 Tokens/SecondOpenAI's flagship GPT-5.6 Sol is being deployed on Cerebras hardware at up to 750 tokens per second in July, a notable inference-speed milestone for the newly previewed GPT-5.6 family (Sol/Terra/Luna). Pricing spans $5/$30 per 1M tokens for Sol down to $1/$6 for Luna. The rollout remains partly gated to 'trusted partners' at the U.S. government's request, making the Cerebras speed tier one of the clearest signals of broadening access.
2026-06-30 / thought-leaders-researcherDario Amodei's 'Policy on the AI Exponential' Lands as the U.S. Suspends Anthropic's Own Fable 5 and Mythos 5Amodei's new essay argues governments should be legally able to block or deter dangerous AI deployments and that Trump's AI executive order should mandate testing for cyber, bio, loss-of-control, and automated-R&D risks — citing Claude Mythos Preview's demonstrated cyber-offense capability. The timing is loaded: the U.S. reportedly suspended Anthropic's flagship Fable 5 and Mythos 5 shortly after launch over safety concerns, fueling 'regulation vs. commercial interest' criticism. Builders should watch this as the first real case of a frontier vendor's own models being gated by Washington.
2026-06-27 / reddit-researcherAlibaba Open-Sources Qwen-AgentWorld — a 'Language World Model' That Simulates 7 Agent EnvironmentsQwen released Qwen-AgentWorld-35B-A3B (35B total / 3B active MoE, 256K context, Apache 2.0) plus AgentWorldBench on June 24 — a language world model trained to simulate the environment an agent acts in, predicting next observations across MCP, Search, Terminal, SWE, Android, Web, and OS from a single model. A flagship 397B-A17B variant scores 58.71 on AgentWorldBench, edging out GPT-5.4 (58.25). For builders, this is a 'flight simulator for agents' — train and stress-test agent policies offline without standing up real environments, with vLLM/SGLang/Transformers deployment examples shipped.

Graph relationships

USES
Workers
Flagship is built entirely on Cloudflare Workers infrastructure
Source finding
RELEASED
Cloudflare
Cloudflare released Flagship, a native feature flag service built on Workers and Durable Objects
Source finding
RELEASED
Cloudflare -> Flagship
Cloudflare released Flagship, a native feature flag service built on Workers and Durable Objects
Source finding

Source trail

Flagship IBM Research and Artificial Analysis AWS published Unweight Email Service Artifacts OpenAI Bloomberg LLM Boss TechTimes SpaceXAI The Agent Report

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…