AI Coding Live News — Real-Time 𝕏/Twitter Updates & Breakthroughs

Microsoft AIMicrosoftAI·Collected 8d ago

👀36

🔥84

Microsoft makes MAI-Image-2.6 and MAI-Image-2.6-Flash generally available in Foundry as MAI-Image-2.7-Preview reaches #3 in Image Edit Arena

Microsoft AI announced on Oct 27 that MAI-Image-2.6 and MAI-Image-2.6-Flash are generally available in Microsoft Foundry, with multi-reference editing from up to five images, Bing web grounding, output up to 1.5K and published per-token pricing. The same day its next image editing model, MAI-Image-2.7-Preview, reached #3 overall in Arena's Image Edit leaderboard at 1470 points, 45 points above 2.6; for now it is available only for editing in MAI Playground.

Key Takeaways

MAI-Image-2.6 and MAI-Image-2.6-Flash are generally available in Microsoft Foundry, and Microsoft says 2.6 is coming soon to Adobe Firefly
Official pricing: 2.6 costs $5 text input, $8 image input and $38 image output per 1M tokens; Flash costs $1.75, $2.50 and $19
2.6 supports multi-reference editing with up to five images, Bing web grounding, dynamic aspect ratios and output up to 1.5K
Per Arena, MAI-Image-2.7-Preview is #3 overall in Image Edit Arena at 1470 points, 45 points above 2.6 (#7), putting Microsoft AI #2 among labs
Per Artificial Analysis, 2.7-Preview is #2 on AA-Image-Editing v2.1 and beats 2.6 on all 7 editing actions, ranking #1 on 4 of them

Read full story

Linked Models & Agents:

⚡ MAI-Image-2.6⚡ MAI-Image-2.6-Flash⚡ MAI-Image-2.7-Preview🌐 Microsoft 🌐 Microsoft AI

#Microsoft AI#MAI-Image#image generation#image editing#Microsoft Foundry#Pricing#Arena

ClaudeDevs@ClaudeDevs·Collected 20d ago

👀94

🔥91

View on 𝕏 ↗

Claude Opus 5.5 fast mode leaves research preview and drops the waitlist: same model, up to 2.5x output speed at $8/$40 per MTok

On 2026-10-15 Anthropic moved Claude Opus 5.5 fast mode out of research preview and removed the waitlist: no access request and no fast-mode-2026-02-01 beta header needed, just speed: "fast". Same weights, up to 2.5x output tokens/sec per Anthropic, at 2x standard price ($8/$40 per MTok). Start/Build/Scale tiers get a default fast rate limit; it is live on Microsoft Foundry (Azure-hosted) with Google Cloud coming soon. Opus 5 and Opus 4.8 fast mode remain in research preview.

Key Takeaways

Pricing: Opus 5.5 fast is $8/$40 per MTok, 2x the $4/$20 standard; Opus 5/4.8 fast stay in preview at $10/$50 (docs)
Speed: up to 2.5x output tokens/sec per Anthropic; gains are OTPS, not TTFT; same weights
Access: no request and no beta header; default dedicated fast limits on Start/Build/Scale tiers (release notes)
Platforms: Claude API (incl. Managed Agents) and Microsoft Foundry Azure-hosted; Google Cloud coming soon; not on Batch API, Priority Tier, or Claude Platform on AWS
Gotchas: switching speeds invalidates the prompt cache; 429 on limit; check usage.speed

Read full story

Linked Models & Agents:

⚡ Claude Opus 5.5 ⚡ Claude Opus 5 ⚡ Claude Opus 4.8 🤖 Claude Managed Agents 🌐 Anthropic 🌐 Microsoft

#Anthropic#Claude Opus 5.5#fast mode#API#Pricing#Microsoft Foundry

Satya Nadella@satyanadella·Collected 26d ago

👀227

🔥88

View on 𝕏 ↗

Microsoft launches Microsoft-Decision-1: a Qwen3.5-9B-based decision-scoring model at $0.042/M input tokens, free output, on Foundry and OpenRouter

On Oct 9 Microsoft released Microsoft-Decision-1, a decision-scoring model that returns calibrated probabilities over fixed options instead of generating text, aimed at routing, classification, verification, agent guardrails and AI judging. Microsoft reports top accuracy across 36 blind benchmarks, ~35x lower P50 latency than GPT-6 Sol and a 1.3% decision-flip rate under perturbation. Available on Microsoft Foundry and OpenRouter, 32K context, $0.042/M input, free output.

Key Takeaways

Pricing: $0.042/M input tokens, $0 output, 32,768-token context (OpenRouter)
Vendor-reported: highest accuracy across 36 blind benchmarks (~150K questions); 2.5x faster than runner-up H2O-Lightning-4B v1.1 and 35x faster than GPT-6 Sol
Robustness: 1.3% average decision flips across 8 perturbation types; zero flips when options are paraphrased, reversed or shuffled
Base: post-trained from Qwen3.5-9B for single-pass scoring; Microsoft plans to rebase on MAI and OpenAI models
Internal use: Xbox Research labeled 10K+ feedback items at near GPT-6 Sol quality, 14x+ faster and ~200x cheaper (vendor-reported)

Read full story

Linked Models & Agents:

⚡ Microsoft-Decision-1⚡ Qwen3.5-9B⚡ GPT-6 Sol 🌐 Microsoft 🌐 OpenRouter

#Decision Models#Microsoft#OpenRouter#Foundry

Satya Nadella@satyanadella·Collected 62d ago

👀56

🔥93

View on 𝕏 ↗

Satya Nadella: GPT-6 Astra live on Azure Microsoft Foundry for early customers

Microsoft CEO Satya Nadella says early customers are already using GPT-6 Astra on Azure, linking to the Microsoft Foundry post that Astra is available for work. It is a concrete availability signal for developers wiring Astra into the Azure enterprise stack.

Key Takeaways

Astra is in Microsoft Foundry / Azure, not only ChatGPT-side chatter.
Nadella confirms early customers are already using it for work workloads.
Developers can follow the Azure blog to wire frontier agent/coding capability in Foundry.

Read full story

Linked Models & Agents:

🌐 OpenAI 🌐 Microsoft Azure AI

#GPT-6 Astra#Microsoft Foundry#Azure#Satya Nadella#Enterprise#OpenAI#API

First Squawk@FirstSquawk·Collected 64d ago

👀57

🔥96

View on 𝕏 ↗

Anthropic launches Claude Fable 5.1; cache reads $0.25, ~25% cheaper typical jobs

Anthropic launched Claude Fable 5.1 on September 1 (API id claude-fable-5-1) for Pro, Max, Team, and Enterprise, plus the Claude API, AWS, Google Cloud, and Microsoft Foundry. List price stays $10 input / $50 output per 1M tokens, but cache reads drop to $0.25 (75% less than Fable 5), which Anthropic estimates cuts typical workloads about 25% and highly agentic ones up to about 45%. Claude Mythos 5.1 remains limited to trusted-access programs.

Key Takeaways

Fable 5.1 is GA for Pro/Max/Team/Enterprise and the API as claude-fable-5-1.
List price stays $10/$50 per 1M tokens; cache reads are $0.25, about 25% cheaper typical jobs and up to ~45% on heavy agent workloads.
Mythos 5.1 stays trusted-access only. Fable 5.1 targets long-horizon coding; Devin said it is moving Opus 5 traffic onto it at launch.

Read full story

Linked Models & Agents:

⚡ Claude 3.7 Sonnet (Hybrid Reasoning)🤖 Claude Code (Anthropic)🤖 Devin 2.0 (Cognition)🌐 Anthropic

#Claude Fable 5.1#Claude Mythos 5.1#Anthropic#Pricing#Cache#Agentic coding#Claude code