Models
The models themselves: families, price per million tokens, code benchmarks, open weights, and the criteria teams use to pick a supplier.
Articles
Market
Microsoft and Google earnings: Azure grows 48%, Google Cloud doubles as AI spending climbs
Azure grew 48% and Google Cloud 103% last quarter as AI capex kept rising. Plus Clover-35B, an open diffusion model, and OpenCode, Zed and Cline updates.
Agents
GitHub Copilot Fusion picks models for you: better coding for less?
Copilot Fusion selects and combines AI models. Its benchmark score, cost per task, limitations and what GitHub credits mean for developers.
Models
Genie 4: Google imagines interactive worlds that move, sound and speak
Genie 4 generates interactive worlds with sound. How Google's model works, what creators might use it for and where it still falls short.
Models
Gemini 4 Carbon: Google retakes the lead but keeps it locked away
Gemini 4 Carbon tops Artificial Analysis and LMArena, but it is not yet available to developers or the general public. Scores, pricing, access.
Models
Open-weight models: who leads, and how far behind closed models
Kimi, DeepSeek, GLM, Qwen, Gemma, gpt-oss: who leads among open-weight models, the gap with closed ones, the licences and what it means for developers.
Market
Oracle opens its cloud to Google TPUs, Meta plans $95B of stock for AI: the October 28 roundup
Oracle will offer Google TPUs in 2027, Meta plans $95B of stock sales for AI, Google ships a CLI for coding agents and Liquid AI releases LFM3.
Market
Anthropic's IPO: the economic strategy behind the S-1
Revenue, a mostly accounting loss, hundreds of billions committed to compute, founders in control: Anthropic's strategy as read in its S-1 filing.
Models
Anthropic files to go public, OpenAI launches GPT-6.1 Terra: the October 27 roundup
Anthropic publishes its S-1 ($16.3B of first-half revenue), OpenAI ships GPT-6.1 Terra at $1/$5 and two deals finance $103B of AI chips.
Coding agents
Anthropic makes the case for a single agent, Nanbeige4.3-33B, Cline Desktop 0.0.49: the October 26 roundup
Anthropic backs a single agent over subagents, Nanbeige ships an open 33B model, Cline Desktop shows diffs and late data centers now cost $5M a day.
Coding agents
OpenAI wins Anthropic customers on price, Block's Dessa, AWS Infinity, Ollama speaks: the October 25 roundup
OpenAI lures Replit and Rippling with discounts, Block unveils Dessa, AWS open-sources the Infinity runtime, Ollama speaks and Qwen Code reaches 0.26.1.
Coding agents
Cursor Computer Use, unified Claude Enterprise, OpenAI trains offline, Agensh: the October 25 roundup
Cursor agents now drive desktop apps, Claude Enterprise merges chat and Cowork, OpenAI trains with no internet access and Microsoft open-sources Agensh.
Models
Kimi K3.1 Pro, Codex CLI 0.164, Copilot in Teams, Replit: the October 24 roundup
Moonshot ships Kimi K3.1 Pro (1M tokens, weights promised by Nov 4), Codex CLI 0.164 adds presets, Copilot lands in Teams and Replit adds agent skills.
Coding agents
Flint runs locally, Replit Monitoring Agent, Google's EGM plugin, Claude on Google Cloud: the October 23 roundup
Recursive ships Flint, a local coding model under Apache 2.0, Replit launches an on-call agent, Google opens its EGM plugin, Claude goes Anthropic-managed.
Coding agents
VS Code 1.143, GPT-6 on Google Cloud, NVIDIA DSX POD: the October 22 roundup
VS Code 1.143 pits several models against one task, Google Cloud adds GPT-6 to Model Garden and NVIDIA promises AI factories set up in 14 days.
Models
AI data centers: $3.7 trillion a year in revenue by 2032 to pay off, according to a paper published by Brookings
A Columbia economist puts US AI investment at $10.3 trillion from 2025 to 2032, and the revenue needed to pay it off at $3.725 trillion a year.
Models
DeepSeek is reportedly helping Huawei design its next AI chips: what is sourced and what is not
According to The Information, Huawei engineers work inside DeepSeek's offices to fit future Ascend chips to its models. The report rests on two anonymous sources.
Models
Gemini 4 Argon: nearly 25 trillion tokens in 72 hours, three times Gemini 3
Google says Gemini 4 Argon processed nearly 25 trillion tokens in the 72 hours after launch, about three times Gemini 3, and that it is short of capacity.
GitHub repos
The most-starred AI GitHub repositories among developers (October 2026)
OpenCode passes 218,000 stars and leads our ranking of 20 AI repositories for developers; coding agents account for 40% of all stars (GitHub API, October 20, 2026).
Practical guides
How to run a coding model locally with Ollama
Install Ollama on macOS, Linux or Windows, pick a model suited to code, set the context window and connect OpenCode or Cline, according to the official documentation.
Models
MiniMax and DeepSeek: Flash models slash the price of agentic coding
Within three days, DeepSeek and then MiniMax shipped "Flash" models that chase frontier coding scores for a few cents per task.
Models
COBOL, Fortran: OpenAI pitches AI for legacy migration
OpenAI reports 86.6% successful COBOL migrations with GPT-6.1 Sol, up from 5.4%. We look at the tests, the real prices and the limits.
Models
Model APIs: Anthropic overtook OpenAI in two years
Between 2023 and 2025, Anthropic went from 12% to 40% of enterprise API spend, OpenAI from 50% to 27%. Open source is shrinking.
Tools
-
Models
Claude
Anthropic
Pricing : Freemium
Anthropic's model family, which grew from 12% to 40% of enterprise LLM API spend between 2023 and 2025 (Menlo Ventures).
- API
-
Models
GPT
OpenAI
Pricing : Paid
OpenAI's models: 27% of enterprise LLM API spend in 2025, down from 50% in 2023 (Menlo Ventures).
- API
-
Models
Open-source models
Various
Pricing : FreeLicense : Varies by model
Open-source share of enterprise API usage fell from 19% to 11% between 2024 and 2025 (Menlo Ventures).
- Open source
-
Models
Muse Spark Open
Meta
Pricing : FreeLicense : Muse Model License
Meta's open-weight MoE model under the Muse licence (not open source), built for coding and agents: 816B parameters (42B active), 1M-token context window (source: Meta, Oct 15, 2026).
- Open weights
- 1M tokens
-
Models
MiniMax M3.1 Flash
MiniMax
Pricing : FreemiumLicense : MiniMax Community License
MiniMax's "Flash" model for agents and coding (Oct 19, 2026): 73.7% on DeepSWE v1.1 and $0.15/$0.60 per million input/output tokens (source: MiniMax).
- API
- Low cost
- Open weights
-
Models
DeepSeek V4.1-Flash-1017
DeepSeek
Pricing : FreemiumLicense : MIT
Flash model update (Oct 17, 2026) with MIT-licensed open weights: 77.0% on DeepSWE v1.1 (source: DeepSeek, Hugging Face).
- Open weights
- API
-
Models
Gemini 4 Argon
Google
Pricing : Paid
Google's frontier model, generally available since Oct 13, 2026; introductory price of $2 input and $10 output per million tokens (source: Google).
- API
- Antigravity