Claude Fable 5.2 (Max, Default Fallback) achieved a score of 65 on the Artificial Analysis Intelligence Index, well above the comparable median of 40. The model supports text and image input with text output and a 1M token context window, but testing showed it used 275M tokens versus a 72.8M median, took 45 minutes per task against a 12.9 minute median, and cost $26.54 per task to evaluate. Input pricing is $8.00 per 1M tokens and output pricing is $40.00 per 1M tokens, both well above typical rates.
artificialanalysis.ai
· 2026-11-09
Meta Superintelligence Lab released Muse Ember, its lightest and cheapest model yet, priced at 10 cents per million input tokens. Meta says the model is built for high-volume agent work like tool calls, coding subtasks, routing, and document parsing, and can process text, images, and audio without needing separate models. Ember is the fourth model in Meta's Muse lineup, following Spark, Glimmer, and Wonder.
techcrunch.com
· 2026-11-09
Muse has released Ember, described as its smallest and most efficient model, built for tool calls, coding subtasks, routing, extraction and long document parsing. According to benchmarking data cited by the company, Ember matches Claude Haiku 5.5's top score at roughly a third of the cost and outperforms GPT-6 Luna at every reasoning setting tested.
research.meta.ai
· 2026-11-09
Thinking Machines Lab, founded by former OpenAI CTO Mira Murati, has launched Inkling 2 Small, a 594-billion-parameter model that accepts images but only outputs text. Third-party evaluator Artificial Analysis gave it a score of 31 on its Intelligence Index, putting it on par with Gemini 3.8 Flash, though the company says it will not release the model's weights publicly until next week.
zdnet.com
· 2026-11-06
The developer unveiled Inkling 2 Small, a Mixture-of-Experts model with 594 billion total parameters and 17 billion active, supporting image inputs and variable effort levels. The model scored 31 on the Artificial Analysis Intelligence Index, a notable jump from the company's prior generation, with gains attributed to more efficient pre-training and a unified reinforcement learning run. It is currently accessible only through the Tinker platform with closed weights, ahead of a planned Apache 2.0 open-weight release next week pending safety testing.
thinkingmachines.ai
· 2026-11-06
Google has launched Gemini 4 Carbon, which Artificial Analysis now ranks as the top AI model on its Intelligence Index V5 suite, narrowly surpassing Anthropic's Claude Opus 5.5 by one point and besting OpenAI's GPT-6 Astra and Meta's Muse Wonder Full. The model also leads a separate Artificial Analysis test measuring hallucination rates and factual recall, though its score of 54 out of a possible 100 shows even the top model remains far from fully reliable. Google is initially rolling the model out to cyber defenders via its Fairwind program, with broader access to follow later.
tomshardware.com
· 2026-11-03
Artificial Analysis reports that Google's new Gemini 4 Carbon model has topped its Intelligence Index v5 with a score of 59, ranking first in 4 of 9 component evaluations including HLE-Diamond and AA-Omniscience. The model also placed second on the firm's Coding Agent Index, trailing only Anthropic's Claude Opus 5.5, and showed sizable gains in agentic knowledge-work benchmarks over its predecessor, Gemini 4 Argon. The model is currently in limited rollout with a 50% launch discount and not yet publicly available.
artificialanalysis.ai
· 2026-10-29
Google has launched Gemini 4 Carbon, its second new frontier AI model in a month following Gemini 4 Argon. The model improves on coding, agentic tasks, scientific reasoning and multimodal understanding, with Google saying its biggest gains are in cybersecurity, including more than doubling secure-code generation rates versus Argon. Carbon will first go to select cyber defenders before wider release, priced at $2 per million input tokens and $10 per million output tokens.
androidauthority.com
· 2026-10-28
A company has unveiled Gemini 4 Carbon, a frontier AI model it says scores highest on the Artificial Analysis Intelligence Index v5.0 at 59, with strong results on cybersecurity benchmarks including SEC-bench Pro, ExploitGym, and CWE-bench. Rather than a full public launch, the model will first go to select cyber defenders this week, then expand to trusted partners via the Fairwind Program, with pricing set at $2 per million input tokens and $10 per million output tokens.
blog.google
· 2026-10-28
Google announced Gemini 4 Carbon, a new AI model it says delivers major improvements in coding, long-running agentic tasks, and cybersecurity defense compared to its predecessor Argon, released earlier this month. The model tops the Artificial Analysis Intelligence Index v5.0 and leads rival models from Anthropic and OpenAI on two security-focused benchmarks, SEC-bench Pro and ExploitGym. Access is currently limited to select cybersecurity defense firms through Google's Fairwind Program, with wider partner access planned later.
theverge.com
· 2026-10-28
Artificial Analysis has released version 5 of its Intelligence Index, the widely cited independent metric ranking frontier AI models, following criticism of flaws in prior benchmark aggregation methods. The piece discusses how AA's index and competitors like Epoch AI's Capabilities Index attempt to distill model performance into a single comparable score, despite inherent trade-offs in how such scores are aggregated.
interconnects.ai
· 2026-10-28
Artificial Analysis launched version 5 of its Intelligence Index, a composite benchmark measuring AI model capabilities, adding two new tests (FrontierCode 2.0 and Terminal-Bench-Science) and upgrading three others, including a refined version of Humanity's Last Exam. The firm also introduced a companion Progress Index to track model capability over time, and reports that 60% of the Index now relies on private or held-out test data. In current results, Claude Opus 5.5 leads with a score of 58, narrowly ahead of GPT-6 Astra at 57.
artificialanalysis.ai
· 2026-10-28
GoKawiil