/
Navigation
C
Chronicles
Browse all articles
C
E
Explore
Semantic exploration
E
R
Research
Entity momentum
R
N
Nexus
Correlations & relationships
N
~
Story Arc
Topic evolution
S
↻
Drift Map
Semantic trajectory animation
D
P
Posts
Analysis & commentary
P
Browse
@
Entities
Companies, people, products, technologies
◇
Domains
Browse by publication source
☉
Handles
Browse by social media handle
Detection
?
Concept Search
Semantic similarity search
!
High Impact Stories
Top coverage by position
+
Sentiment Analysis
Positive/negative coverage
*
Anomaly Detection
Unusual coverage patterns
Analysis
vs
Rivalry Report
Compare two entities head-to-head
/\
Semantic Pivots
Narrative discontinuities
!!
Crisis Response
Event recovery patterns
Connected
Nav: C E R N
Search: /
Command: ⌘K
Embeddings: large
VOICE ARCHIVE

Skyler Miao

@skylermiao7
7 posts
2026-02-13
my favorite part about M2.5: SOTA without the speed tax. SWE-bench 74 → 80.2, and the best thing — 30% faster. no trade-off.
2026-02-13 View on X
MiniMax

MiniMax releases M2.5, claiming the model delivers on the “intelligence too cheap to meter” promise, priced at $0.30/1M input tokens and $1.20/1M output tokens

Today we're introducing our latest model, MiniMax-M2.5.  —  Extensively trained with reinforcement learning …

View original
With M2.5 we're also shipping MiniMax Experts. general agents sound nice until you realize they don't know your stack, your domain, or how you actually work. now you can build experts that do. share yours, grab others'. community already made: clawbot assistant, crypto trading
2026-02-13 View on X
MiniMax

MiniMax releases M2.5, claiming the model delivers on the “intelligence too cheap to meter” promise, priced at $0.30/1M input tokens and $1.20/1M output tokens

Today we're introducing our latest model, MiniMax-M2.5.  —  Extensively trained with reinforcement learning …

View original
2026-02-12
my favorite part about M2.5: SOTA without the speed tax. SWE-bench 74 → 80.2, and the best thing — 30% faster. no trade-off.
2026-02-12 View on X
MiniMax

MiniMax releases M2.5, claiming the model delivers on the “intelligence too cheap to meter” promise, priced at $0.30/1M input tokens and $1.20/1M output tokens

Today we're introducing our latest model, MiniMax-M2.5.  —  Extensively trained with reinforcement learning …

View original
With M2.5 we're also shipping MiniMax Experts. general agents sound nice until you realize they don't know your stack, your domain, or how you actually work. now you can build experts that do. share yours, grab others'. community already made: clawbot assistant, crypto trading
2026-02-12 View on X
MiniMax

MiniMax releases M2.5, claiming the model delivers on the “intelligence too cheap to meter” promise, priced at $0.30/1M input tokens and $1.20/1M output tokens

Today we're introducing our latest model, MiniMax-M2.5.  —  Extensively trained with reinforcement learning …

View original
2026-01-10
Minimax M2.x, M3, and our Coding Plan will stay open to third-party apps and all users. We believe the vitality, creativity, and innovative of the community—just look at what apps like OpenCode, Kilo, Roo, and Cline are achieving. Open Models + Open Apps delivered great user
2026-01-10 View on X
VentureBeat

Anthropic adds safeguards to prevent third-party apps, like OpenCode, from spoofing Claude Code to access Claude models for more favorable pricing and limits

Anthropic has confirmed the implementation of strict new technical safeguards preventing third-party applications from spoofing …

View original
Congrats to OpenCode and the community. If Bun ever got cut off too, that would be genuinely shocking. Rooting for the open-source community, happy to support however we can.
2026-01-10 View on X
VentureBeat

Anthropic adds safeguards to prevent third-party apps, like OpenCode, from spoofing Claude Code to access Claude models for more favorable pricing and limits

Anthropic has confirmed the implementation of strict new technical safeguards preventing third-party applications from spoofing …

View original
2025-11-07
Congrats! Great to see an OSS model outperforming closed ones on HLE, and another model adopting interleaved thinking with impressive results. Really cool to see our efforts with partners, like @OpenRouterAI and @cline, to support interleaved thinking now benefiting more users
2025-11-07 View on X
CNBC

Chinese startup Moonshot releases Kimi K2 Thinking, an open-weight model it claims beats GPT-5 in agentic capabilities; source: the model cost $4.6M to train

Chinese startup Moonshot on Thursday released its latest generative artificial intelligence model which claims to beat OpenAI's ChatGPT in …

View original