Tests & comparisons
xAI
Alex
xAIAlexAlex’s AI posts
Alex Getman’s posts on AI models, tools, releases, prices and usage limits, newest first.
SearchMon 21 Sep
ReleasesxAIAlex
xAIAlexGrok 4.7 is out!
Leaks
OpenAIAlex
OpenAIAlexGPT-6 Luna may already be in A/B testing
Some GPT-5.6 Luna requests are being served as GPT-6 Luna behind the scenes.
LeaksOpenAIAlex
OpenAIAlexTibo hinted that a big ship is happening today
My guess is GPT-6 Sol, and we’ll probably get a Codex reset too.
Sun 20 Sep
Leaks
AnthropicAlex
AnthropicAlexNew Claude Opus drops early next week
It will likely be called Opus 5.5 $4/$20 input/output pricing, 20% cheaper than Opus 5
Tests & comparisonsAnthropicAlex
AnthropicAlexFinished my ClaudeCut experiment
I wanted to see how much of a $20 Claude subscription actually goes into the work, and how much gets eaten by Claude Code itself. I cut it down to…
Sat 19 Sep
Tools & updatesAnthropicAlex
AnthropicAlexI castrated Claude Code
Ripped almost everything out and left only the bare minimum: read, inspect, write. Startup context is now under 2k tokens. I also capped the…
Fri 18 Sep
Tools & updatesAnthropicAlex
AnthropicAlexClaude added support for AGENTS.md
Tests & comparisons
Z.aiAlex
Z.aiAlexZCode uploads full project snapshots to the cloud
A researcher noticed a huge ZCode folder and started digging into it. Inside was a 313 MB encrypted copy of his work project. He then reverse…
Tools & updates
MiniMaxAlex
MiniMaxAlexMiniMax open sourced MiniMax Code CLI
It runs on a heavily modified Pi stack with MiniMax’s own agent runtime and tooling. In FrontierHarness Eval, MiniMax Code hit 76.7% pass rate, 4m…
Tests & comparisonsHugging FaceAlex
AlexResearchers hacked OpenAI through a malicious image
They found a bug in how OpenAI’s forum processed HEIC images. Uploading a crafted image gave them control over the forum server. From there, a flaw…
ReleasesZ.aiAlex
Z.aiAlexZ.ai launches GLM 5.3 FlashX at up to 200 tokens/s
It is essentially the same 320B GLM 5.3 Flash, but with a heavily optimized inference stack and up to 5× faster generation. Runs on 100,000+ Chinese…
Tests & comparisonsOpenAIAlex
OpenAIAlexAn unreleased Astra model started prompt injecting its own context during RL training
Leaks
GoogleAlex
GoogleAlexGemini 4.0 Pro specs leaked
• Pricing: $2.25/$11.25 • 2M token context window • Beats GPT-6 Astra and Claude Fable 5.1 on almost every reported benchmark • While being…
Thu 17 Sep
Releases
AlibabaAlex
AlibabaAlexQwen3.8 Omni Flash is out
It can see, hear and understand video, then actually act on it Give it hours of footage and it can find the right moments, translate, edit or turn…
LeaksxAIAlex
xAIAlexGrok 4.7 just appeared in Google Cloud
That usually means the official release is very close.
LeaksGoogleAlex
GoogleAlexGemini 3.8 Flash just appeared on Arena again
Looks like Google is testing a fresh checkpoint under that name. My bet: either Gemini 3.9 Flash or an early Gemini 4.0 Pro checkpoint.
Prices & limitsAnthropicAlex
AnthropicAlexMy Claude weekly usage was at like 90% and then suddenly dropped to 60%
Guess Anthropic gave a bunch of us some compensation? Did this happen to anyone else?
Wed 16 Sep
LeaksAlex
AlexUnion Alpha just dropped in stealth on OpenCode
It scores around 51% on Terminal Bench v4 at roughly $1.5 per task, putting it close to GPT 6 Astra high for much less. My guess is this is a new…
ReleasesTypeSafe AIAlex
AlexJev is out, a new AI model from startup TypeSafe AI
It is built for fast decision making inside apps. You give it context and possible actions, and it picks the best option with probability and…
Tue 15 Sep
LeaksOpenAIAlex
OpenAIAlexGPT 6 Sol and Luna could launch this week
Terra may be gone for good. The model never found enough demand or a clear price point. Sol could get a bit cheaper, while Luna could move up in…
LeaksAnthropicAlex
AnthropicAlexOpus 5.2 appeared in the Microsoft Foundry catalog
At first glance it looks like a leak, but the -2 may simply refer to Foundry’s version 2 deployment rather than a new Claude model. Still worth…
Mon 14 Sep
Sat 12 Sep
Prices & limitsOpenAIAlex
OpenAIAlexCODEX LIMITS RESET
Thu 10 Sep
Prices & limitsOpenAIAlex
OpenAIAlexOpenAI is pausing sales of the ChatGPT Pro x20 subscription
Releases
DeepSeekAlex
DeepSeekAlexDeepSeek V4.1 Flash is out
• New architecture: it only uses the part of the model it actually needs, making it faster and cheaper. • New KV cache: long conversations and agent…
Wed 9 Sep
Prices & limitsOpenAIAlex
OpenAIAlexReddit users say Astra xHigh lasts much longer than Medium
One Pro 20x user ran xHigh for an hour and used only 2% of their weekly usage, while Medium drained it much faster, and others said they saw the…
Prices & limitsOpenAIAlex
OpenAIAlexThinking of upgrading to ChatGPT Pro?
Might be your last chance. Codex demand is growing so fast that new Pro subscriptions could be paused.
Tue 8 Sep
ReleasesOpenAIAlex
OpenAIAlexChatGPT Images 2.5 is out
Up to 50% faster generation - Much better reference photo fidelity and more precise edits - Sketch turns a rough drawing into a finished image
Tests & comparisonsOpenAIAlex
OpenAIAlexOpenAI has a new internal model that is way ahead of GPT 6 Astra in math
On a curated set of open math problems, it reaches about 48% pass rate with more test time compute. GPT 6 Astra peaks at around 17%.