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

Vercel Developersvercel_dev·Collected 2d ago

👀17

🔥83

Vercel AI Gateway Adds DeepSeek V4.1 Flash 1017 Ultrafast: Vendor Figures Cite Upwards of 1,000 Tokens per Second on a Fireworks Endpoint

Vercel published a changelog on 2026-11-07 (authors Jerilyn Zheng and Zachary Chen) saying DeepSeek V4.1 Flash 1017 Ultrafast is now on AI Gateway through Fireworks' accelerated serverless endpoint; the official @vercel_dev account posted at 17:25 UTC. Vendor figures say it can generate upwards of 1,000 tokens per second, serving the 2026-10-17 V4.1 Flash update rather than new base-model weights. The model page lists a 1,048,576-token context window and pricing of $2.40 per million input tokens and $9.60 per million output tokens; it is a public preview with no waitroom or session time limit. The model ID is deepseek/deepseek-v4.1-flash-1017-ultrafast, callable via the AI SDK, an OpenAI-compatible API, and npx vercel ai-gateway setup for Claude Code, Codex and Cursor.

Key Takeaways

Release: Vercel changelog dated 2026-11-07 by Jerilyn Zheng and Zachary Chen; @vercel_dev posted at 17:25 UTC (about 909 likes, 390 bookmarks and about 98K impressions at capture)
Model ID: deepseek/deepseek-v4.1-flash-1017-ultrafast, serving the 2026-10-17 V4.1 Flash update on Fireworks' accelerated serverless endpoint, not new base-model weights
Speed (vendor figures): upwards of 1,000 tokens per second; the model page lists both the context window and max output tokens as 1,048,576; public preview with no waitroom or session time limit
Pricing (Vercel model page): $2.40 per million input tokens and $9.60 per million output tokens; the changelog says AI Gateway reflects provider pricing with no markup and no platform fee on BYOK inference
Access: AI SDK or base_url https://ai-gateway.vercel.sh/v1; Claude Code, Codex and Cursor can use npx vercel ai-gateway setup; fx can run /model deepseek/deepseek-v4.1-flash-1017-ultrafast

Read full story

Linked Models & Agents:

⚡ DeepSeek-V4.1-Flash 🤖 Claude Code (Anthropic)🌐 DeepSeek

#Vercel#AI Gateway#DeepSeek V4.1 Flash#Ultrafast#Fireworks#Pricing

Tibothsottiaux·Collected 2d ago

👀54

🔥82

View on 𝕏 ↗

X Discussion: OpenAI's Tibo Asks "Would a Pro 50 Make Sense?" After Saying Pro 100/200/500 Give 10x/20x/50x Plus Usage With No 5-Hour Limits

Tibo (@thsottiaux), who works on Codex and ChatGPT at OpenAI, clarified on X at 05:09 UTC on 2026-11-08 that all Pro plans have no 5-hour limits and give, compared with Plus, about 10x usage on Pro 100, 20x on Pro 200 and 50x on Pro 500, which also includes Ultrafast. Thirty-six minutes later he quoted that post and asked, "Would a Pro 50 make sense? Wdyt?" At capture the question had about 4,390 likes, 2,013 replies, 269 quotes and about 438,600 impressions. These are an employee's personal posts and a request for feedback, not a new plan OpenAI has launched. The official Help Center still lists only Pro 100, Pro 200 and Pro 500 and does not state exact multipliers.

Key Takeaways

Question post: Tibo (@thsottiaux) wrote "Would a Pro 50 make sense? Wdyt?" at 05:45 UTC on 2026-11-08; about 4,390 likes, 2,013 replies, 126 reposts, 269 quotes and about 438,618 impressions at capture
Quoted clarification (05:09 UTC): no Pro plan has 5-hour limits; versus Plus, Pro 100 gives about 10x, Pro 200 about 20x and Pro 500 about 50x usage plus Ultrafast; about 4,924 likes, 1,003 replies and about 873,738 impressions at capture
Official wording: the OpenAI Help Center lists three Pro plans at $100, $200 and $500 per month; Ultrafast is only on Pro 500 and uses the included allowance at 8x the Standard rate; the page does not state the 10x/20x/50x figures
Discussion: many quote posts back a tier between Plus and Pro 100, while others ask for higher Plus limits or a new model first; OpenAI has not announced any Pro 50 price, multiplier or feature set
Framing: this is a discussion started by an employee's personal posts, not a launched plan

Read full story

Linked Models & Agents:

🤖 openai-codex 🌐 OpenAI

#X discussion#Tibo#Thsottiaux#ChatGPT Pro#Pro 50#Codex#usage limits#OpenAI

Vivix@AICoder·Collected 30d ago

👀166

🔥88

Vivix launches W1, a streaming-native real-time video model: ~30B active params, $0.003/sec promo, API in preview

Vivix released Vivix-W1, a streaming-native multimodal video model that accepts text, image, audio and video references and lets new inputs steer the stream mid-generation. Vivix says it has ~30B active parameters and costs $0.003 per output second through Oct 31; in vendor tests an 8s 720p reference-to-video job took 8.4s on average. The API is in limited preview with standard, fast and ultrafast tiers.

Key Takeaways

Vendor-stated ~30B active parameters with native audio-video, multi-shot and continuous streaming
$0.003/sec promo through Oct 31 vs vendor-listed $0.080 (H3 Max), $0.112 (Kling 3.0), $0.231 (Seedance 2.5)
Vendor test: 8s 720p ref-to-video in 8.4s avg vs 145.9s Kling 3.0 and 272.4s Seedance 2.5 (excl. queue)
API tiers vivix-w1 / fast / ultrafast at 1/3/4 credits per second, 5–15s clips
No third-party quality benchmark yet

Read full story

Linked Models & Agents:

⚡ Vivix-W1⚡ Kling 3.0⚡ Seedance 2.5⚡ MiniMax H3 Max🌐 Vivix

#API

OpenAI DevelopersOpenAIDevs·Collected 32d ago

👀203

🔥94

View on 𝕏 ↗

OpenAI rolls out GPT-6.1 Sol Ultrafast across API, Codex, and ChatGPT Work: near-Astra intelligence, up to 8x Sol Standard, $12/$60 per 1M tokens

On Oct 8 OpenAI Developers said Ultrafast for GPT-6.1 Sol is rolling out in the API, Codex, and ChatGPT Work. Docs confirm service_tier ultrafast at 6x Standard ($12/$60 short-context), with US/EU residency; Codex/Work access on Pro $500 and eligible Enterprise/Edu. Codex also shipped instant steering the same day.

Key Takeaways

Surface: API + Codex + ChatGPT Work same day; OpenAIDevs post ~199K views / 2,398 likes
Price: Ultrafast = 6x Standard → $12/M in / $60/M out short-context ($0.60 cached, $15 cache write); long-context 2x those
Call: model gpt-6.1-sol with service_tier ultrafast; WebSockets recommended for tool-heavy agents
Access: Codex/Work on Pro $500, eligible Enterprise, credit Edu; Enterprise off by default; US/EU residency
Companion: Codex Day 4 instant steering for realtime course-correction alongside Ultrafast

Read full story

Linked Models & Agents:

⚡ GPT-6.1 Sol ⚡ GPT-6 Astra 🤖 Codex 🤖 ChatGPT Work 🌐 OpenAI

#OpenAI#GPT-6.1 Sol#Ultrafast#Codex#API

Vercelvercel·Collected 40d ago

👀71

🔥90

Vercel AI SDK: [email protected]–7.0.126 speech telemetry + tool-approval clear; @ai-sdk/openai ultrafast service tier

Vercel shipped [email protected] (speech/transcription telemetry) through [email protected] (clear tool approvals on addToolOutput) on 2026-09-30—22 commits since covered [email protected]. Same day @ai-sdk/[email protected] adds OpenAI serviceTier: 'ultrafast' (docs: access-controlled, gpt-5.6-sol), plus Azure MAI speech/transcribe and MCP OAuth conditional invalidation patches in the tree.

Vercel AI SDK: ai@7.0.124–7.0.126 speech telemetry + tool-approval clear; @ai-sdk/openai ultrafast service tier

Key Takeaways

[email protected] (17:48:28Z) → [email protected] (22:57:07Z); 22 commits since 7.0.123
Speech/transcription telemetry + provider usage + streaming lifecycle callbacks (8c65988)
Clear tool approvals when addToolOutput runs (4f3d236); convertDataPart / multipart tool results in tree
@ai-sdk/[email protected]: serviceTier 'ultrafast' (docs: gpt-5.6-sol access-controlled)
npm i [email protected] @ai-sdk/[email protected]; https://ai-sdk.dev/docs/introduction

Read full story

Linked Models & Agents:

⚡ GPT-5.6 Sol 🌐 vercel 🌐 OpenAI

#Vercel AI SDK#[email protected]#Speech#Telemetry#Ultrafast#Tool approval#OpenAI#Release

OpenAI@OpenAI·Collected 42d ago

👀80

🔥93

OpenAI Python SDK 3.20.0: Agents credential/session options, Cyber access programs on Responses, opt-in WebSocket text/tool snapshots

OpenAI shipped Python SDK v3.20.0 on 2026-09-28: Agents credentials gain metadata and update-without-replace auth, sessions add ultrafast tier plus misalignment_policy_violation; Responses can select Cyber access programs (standard/daybreak_blue/daybreak_red); and an opt-in ResponsesWebSocketAccumulator yields immutable incremental text/tool snapshots. Live/Realtime WebSocket query/queue/TLS retry fixes land in the same cut.

Key Takeaways

Shipped v3.20.0 — Agents credential metadata/session ultrafast + Cyber access programs
Opt-in ResponsesWebSocketAccumulator for incremental text/tool snapshots (190+ WS tests)
Live/Realtime WebSocket query/queue/TLS retry hardening
Upgrade: pip install -U openai

Read full story

Linked Models & Agents:

⚡ GPT-6 Sol ⚡ GPT-6 Luna 🤖 OpenAI-Agents 🌐 OpenAI

#OpenAI#Python SDK#Agents#Responses API#WebSocket#Cyber#Realtime

OpenAI@OpenAI·Collected 69d ago

👀53

🔥89

View on 𝕏 ↗

OpenAI Previews Ultrafast Mode: GPT-5.6 Sol at up to 14x Speed

OpenAI previewed Ultrafast mode for GPT-5.6 Sol, powered by Cerebras, generating up to 750 tokens per second—as much as 14x faster. It launches first in the OpenAI API for a select group of customers, targeting latency-sensitive workflows such as real-time voice, commerce, coding, financial research, and security response.

Key Takeaways

GPT-5.6 Sol Ultrafast hits up to 750 tokens/sec, about 14x faster
Inference is powered by Cerebras for latency-critical products
Rolling out first via the OpenAI API for coding, voice, commerce, and security work

Read full story

Linked Models & Agents:

⚡ GPT-5.6 Sol 🌐 OpenAI 🌐 Cerebras

#AI#LLM#Coding#OpenAI#Inference

OpenAI@OpenAI·Collected 71d ago

👀49

🔥93

View on 𝕏 ↗

Ultrafast GPT-5.6 Sol preview: up to 14x speed and 750 tokens/sec for coding and realtime work

OpenAI previewed Ultrafast mode: GPT-5.6 Sol at up to 14x speed, powered by Cerebras, generating up to 750 tokens per second. It launches first in the OpenAI API for a select customer group, then expands as capacity grows. Target workflows include realtime voice and support, commerce, coding and design, financial research, and security response—frontier intelligence where every second counts.

Key Takeaways

Frontier coding quality at near-instant completion speed is now a product mode, not just a benchmark demo.
750 tokens/sec is Cerebras-backed; capacity, not the model card, is the rollout bottleneck.
Access is invite/select-customer preview, so production adoption still waits on fleet growth.

Read full story

Linked Models & Agents:

⚡ GPT-4o 🤖 copilot 🌐 OpenAI

#Ultrafast#gpt-5.6-sol#Cerebras#Latency#Coding#API

OpenAI@OpenAI·Collected 71d ago

👀49

🔥96

View on 𝕏 ↗

OpenAI Previews Ultrafast Mode: GPT-5.6 Sol Generates Up to 750 Tokens/sec Powered by Cerebras

OpenAI previewed Ultrafast mode for GPT-5.6 Sol powered by Cerebras, achieving up to 750 tokens per second (14x standard speed) targeted at real-time voice, low-latency agents, and high-throughput coding workflows.

Key Takeaways

Powered by Cerebras wafer-scale chips, GPT-5.6 Sol reaches up to 750 tokens/s (~14x speedup);
Rolling out to select API customers targeting high-throughput coding, instant code refactoring, and voice agents;
End-to-end full stack optimization designed specifically for latency-critical developer workflows.

Read full story

Linked Models & Agents:

⚡ GPT-5.6 Sol⚡ OpenAI GPT-5 Codex

OpenAI@OpenAI·Collected 71d ago

👀56

🔥94

View on 𝕏 ↗

OpenAI previews Ultrafast mode: GPT-5.6 Sol runs up to 14x faster at 750 tokens/s

OpenAI previewed Ultrafast mode in the API for select developer teams. Powered by Cerebras hardware, it delivers up to ~750 tokens/sec (up to 14x faster generation). Target use cases include realtime voice, high-throughput coding and design, and automated security incident response.

Key Takeaways

Powered by Cerebras inference hardware, output throughput peaks at ~750 tokens/s.
Brings frontier-class intelligence into latency-critical real-time coding workflows.
Coding, design generation, and security response are primary target workloads.

Read full story

Linked Models & Agents:

⚡ GPT-4o ⚡ OpenAI o3-mini (High Reasoning Coding)🌐 OpenAI

#OpenAI#Gpt 5 6 sol#Ultrafast#Cerebras#Inference speed#Realtime coding

OpenAI@OpenAI·Collected 71d ago

👀50

🔥93

View on 𝕏 ↗

OpenAI previews Ultrafast mode: GPT-5.6 Sol at up to 14x the speed

OpenAI previewed Ultrafast mode, saying GPT-5.6 Sol can run at up to 14x its usual speed. Powered by Cerebras, Ultrafast generates up to 750 tokens per second for workflows where latency is the product: real-time voice and support, commerce, coding and design, financial research, and security response. It launches first in the OpenAI API to a select group of customers, with broader business access as capacity grows.

Key Takeaways

Flagship coding model Sol is now being sold on near-real-time throughput, not just quality.
750 tokens/second makes interactive coding agents, voice, and support workflows viable.
Access is still a limited API preview; capacity is the gate to broader rollout.

Read full story

Linked Models & Agents:

⚡ GPT-5.6 Sol 🌐 OpenAI

#Gpt 5 6 sol#Ultrafast#Cerebras#Latency#Codex#API