Thinking Machines Inkling

“Closest in size that's better across that range of benchmarks is Inkling-Small I think, at 276B params”

“yet another mediocre but fully-made-in-the-West open model”

“Every doc opens in a browser straight from disk. No server, no build step, no dependencies.”

“all are fairly decent options at a variety of hardware costs”

“Inkling, e.g. is more accurate in not telling you to adjust valves that are simply not adjustable”

“their Inkling model was decent (admittedly not great though)”

“it's an American open model and is really good!”

“the only big other open source US model that was worth looking at”

“it is SUPER FAST and does a good job at being a terminal buddy”

“Just saw Thinking Machines Inkling model succeed in certain queries”

“Inkling is the most exciting model launch we've seen from an American lab in a while”

“inkling on its own looks high-quality, but expecting companies to spend $ figuring it out before spending more $ on the actual fine-tuning job seems ... hard”

“first seriously competitive non-Chinese open model in a good bit now”

“I enjoyed turning off web search for Inkling to experiment with what innate knowledge is encoded in the model weights.”

“artificial analysis puts it quite low below previous gen open weights models from China”

“it debuts at #41 on the AA index and the post itself calls it not the strongest overall”

“It is surprisingly capable for coding despite its tiny size.”

“It looks promising, but on my cases it still fell short of GPT-5.4 and GPT-5.6 Luna.”

“not obsessed with benchmaxxing on coding but rather being adaptable to the task required”

“I got far enough to run some tests on the openai endpoint, albeit with some finagling”

“accurately answered a number of questions focused on very niche details”

“open source matches their business model, which is to provide fine tuning services for enterprises, etc.”

“It does worse in most benchmarks”

“This is the best voice/tone I've seen from any model so far.”

“This is the first competitive non-Chinese open weights model since what, Llama 3?”

“I always have to clear context btw 150k-400k context even if the context window is 1M on paper.”

“accurate, inexpensive open-weight workhorse for high-level prioritization and decision-making work”

“seems pretty dang snappy and I like it's tone/personality so far”

“This seems like a really really great debut model for a new lab”

“Very impressive model, exciting to see an American open-source lab with such competitive results.”

“Excited to see the next generations of Thinky models.”

“this model should really shine in Autoresearch NanoGPT-style speedruns”

“releasing a decent open weights model in 16 months is actually quite comparable”

“This seems like an extremely strong model for a first release”

“For a first model, and given it's open, I am gaining some faith in American Open research labs again...”

“debuts at 41 on the Artificial Analysis Intelligence Index”

“This has all the right ingredients for success.”

“Its not as good as GLM 5.2 for agentic workflows while also being bigger.”

“Seems like this is particularly good at instruction following”

“best western open weights model- very cool.”