







Muse Ember costs less per task than similarly priced modelsMuse Ember (max) has the same input and output token prices as Claude Haiku 5.5 (max) and GPT-6 Luna (max), but its cached input, most of the cost in our agentic evaluations, is 60% cheaper. Muse Ember (max) also uses… pic.twitter.com/uog9Gl8vsZ— Artificial Analysis (@ArtificialAnlys) November 9, 2026
Artificial Analysis on Twitter / X
Meta's newest model Muse Ember scores 36 on the Artificial Analysis Intelligence Index, matching Claude Haiku 5.5 (max) at a third of the Cost per TaskMuse Ember is Meta's newest efficient-tier model line, with the same input and output token prices as Claude Haiku 5.5 and… pic.twitter.com/ZXq4Zw8DZj— Artificial Analysis (@ArtificialAnlys) November 9, 2026
Muse Spark 1.3 (max) - Intelligence, Performance & Price Analysis | Artificial Analysis
Analysis of Meta's Muse Spark 1.3 (max) and comparison to other AI models across key metrics including quality, price, performance (tokens per second & time to first token), context window & more.
Introducing Our Smallest, Most Efficient Model: Muse Ember
Muse Ember is our smallest and most efficient model, designed for complex reasoning execution and the work agents repeat all day.

Muse Spark 1.3 (xhigh) - Intelligence, Performance & Price Analysis | Artificial Analysis
Analysis of Meta's Muse Spark 1.3 (xhigh) and comparison to other AI models across key metrics including quality, price, performance (tokens per second & time to first token), context window & more.
Artificial Analysis on Twitter / X
We benchmarked Apple's new On-Device model: trails most Gemma and Qwen on-device suitable models but still very usefulGPQA Diamond performance trailed models that are suitable for on-device use such as the smaller Gemma models (3n E4B, 4B, 12B) and Qwen3 models (1.7B, 4B, 8B).… pic.twitter.com/wMrNM7yinL— Artificial Analysis (@ArtificialAnlys) June 20, 2025


Muse Creates Detailed Profiles of All Your Friends and Family
Millions have downloaded Meta’s AI agent Muse. But getting it to do your bidding comes with privacy costs.

tokens are getting more expensive
"language models will get cheaper by 10x" will not save ai subscriptions from the short squeeze

Artificial Analysis v5 and the benchmarker's dilemma
A sort of review of the benchmarking industry and what's missing.

The Ads Model for Prompts Vertically Integrates AI
Meta's dual-tier AI pricing introduces an explicit barter : a 92% discount on inference in exchange for your data. Applying Michael Mauboussin's insight that prices contain information reveals the clearing price of prompt data : $1.24 per million tokens.
A broken pricing paradigm
A token is not a fixed unit of cost Variance in usage creates an interconnected pricing and scaling issue Anjali Shrivastava anjali.shrivastava99@gmail.com | @anjali_shriva THIS ESSAY IS NOW LIVE! A token is not a fixed unit of cost PART I: High variance in AI demand breaks unit economics and re...
Elon Musk appears to have throttled Grok Bot for being “unsustainably” expensive.
Musk hinted at the change in an X post after users complained they were hitting usage limits sooner than expected. The cost of running xAI’s always-on AI agent was “unsustainably high relative to revenue,” he said, without providing details on what has changed. [Media: https://twitter.com/elonmusk/status/2116992258520064397?s=20]

Kimi K2 0711 - API Pricing & Benchmarks
Kimi K2 Instruct is a large-scale Mixture-of-Experts (MoE) language model developed by Moonshot AI, featuring 1 trillion total parameters with 32 billion active per forward pass. $0.57 per million input tokens, $2.30 per million output tokens. 131,072 token context window, maximum output of 32,768 tokens. Includes independent benchmarks from Artificial Analysis.
Announcing VibeBench: The AI benchmark that measures what matters — how models like Opus-4.7 actually feel to use in real-world work.
My coworkers and I have been long-time users of Claude Code and Codex and are getting a ton of exposure to other models due to our deep dives into…
Jev: Fable-level model that doesn’t charge for output tokens because they’re too cheap to meter it’s not general though, it only makes decisions, doesn’t generate text, but input tokens are measured by the billion ($42/btok) typesafe.ai/blog/introducing-system-one-m…

AI 'surveillance pricing' under scrutiny from U.S. lawmakers

Canada Has a Precedent to Regulate Surveillance Pricing
Instacart’s AI-Enabled Pricing Experiments May Be Inflating Your Grocery Bill, CR and Groundwork Collaborative Investigation Finds via @ConsumerReports
Kroger Stores Overcharging Shoppers on Sale Items, CR Price Check Finds via @ConsumerReports
Different Prices for the Same Ride: How Uber and Lyft Use AI to Get More Money Out of You via @ConsumerReports