GLM-4.7-Flash
Hacker ApprovedRanked 17 of 39 open-weight language models by recommenders
Recommended as a fast, capable local model for coding and tool calling, roughly on par with mid-size rivals. Critics find it weaker than bigger models and report chat template bugs.
Recommended as a fast, capable local model for coding and tool calling, roughly on par with mid-size rivals. Critics find it weaker than bigger models and report chat template bugs.
57 people gave an opinion
- recommend it
- 46
- criticize it
- 11
- approval share
- 81%
Recommended in 7 of 13 months
- November 2025: not recommended
- December 2025: not recommended
- January 2026: recommended
- February 2026: recommended
- March 2026: recommended
- April 2026: recommended
- May 2026: recommended
- June 2026: recommended
- July 2026: recommended
- August 2026: not recommended
- September 2026: not recommended
- October 2026: not recommended
- November 2026: not recommended (month in progress)
Why it carries the badge
- 46 people recommend itNeeds 10 or more, met.
- 4.2 recommenders for every criticNeeds 2 to 1 or better, met.
- Recommended in 7 of 13 monthsNeeds 4 or more, met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Strong tool calling and agentic coding
“GLM 4.7 Flash is the main model I use locally for agentic coding work, it's pretty incredible.”
Comment from 7 Apr 2026 in “GLM-5.1: Towards Long-Horizon Tasks” “Community reports highlight strong performance in UI generation and tool calling”
Comment from 20 Jan 2026 in “GLM-4.7-Flash: 30B MoE model achieves 59.2% on SWE-bench, runs on 24GB GPUs” “it actually worked nice in opencode, which I couldn't get Nemotron to do”
Comment from 20 Jan 2026 in “GLM-4.7-Flash” - Comment from 3 Apr 2026 in “April 2026 TLDR Setup for Ollama and Gemma 4 26B on a Mac mini”
Competes with larger models
“it seems to be roughly on par with o3/GPT-5 mini, which would be pretty impressive if it translated to real-world usage”
Comment from 19 Jan 2026 in “GLM-4.7-Flash” “it seems to be roughly on par with o3/GPT-5 mini, which is pretty impressive”
Comment from 19 Jan 2026 in “GLM-4.7-Flash” “in that space, 4.7 Flash holds its own better than it has any right to.”
Comment from 12 Feb 2026 in “GLM-5: From Vibe Coding to Agentic Engineering” - Comment from 19 Jan 2026 in “GLM-4.7-Flash”
Fast on modest hardware
“GLM 4.7 flash is also incredibly performant on modest hardware”
Comment from 4 Mar 2026 in “Something is afoot in the land of Qwen” “Only model that is competitive in speed is GLM 4.7 flash”
Comment from 4 Mar 2026 in “Something is afoot in the land of Qwen” “And you can run quantized versions on old hardware!”
Comment from 12 Feb 2026 in “GLM-5: From Vibe Coding to Agentic Engineering” - Comment from 26 Jan 2026 in “Qwen3-Max-Thinking”
Good local model choice
“glm 4.7 flash which is really good (to be local)”
Comment from 1 Mar 2026 in “Qwen3.5 122B and 35B models offer Sonnet 4.5 performance on local computers” “glm 4.7-flash that was my local model of choice”
Comment from 3 Feb 2026 in “Qwen3-Coder-Next” “is a nice local model for this sort of thing”
Comment from 4 Feb 2026 in “Claude Code: connect to a local model when your quota runs out” - Comment from 13 Jun 2026 in “GLM 5.2 Is Out”
What people warn about
Each mark is one comment
Weaker than bigger models
“it is nowhere near as capable as its bigger sibling in use cases beyond typical chat”
Comment from 16 Feb 2026 in “Qwen3.5: Towards Native Multimodal Agents” “quite fast on the 6000, at around 100 t/s, but definitely not as smart”
Comment from 28 Feb 2026 in “Qwen3.5 122B and 35B models offer Sonnet 4.5 performance on local computers” “it feels dramatically worse than gpt-oss-20b”
Comment from 19 Jan 2026 in “GLM-4.7-Flash” Chat template and implementation bugs
“Seems to work okay, but there usually are subtle bugs in the implementation or chat template”
Comment from 19 Jan 2026 in “GLM-4.7-Flash” “It's really fast! But, for now it outputs garbage because there is no (good) template.”
Comment from 19 Jan 2026 in “GLM-4.7-Flash” Poor experience, failed tool calls
“I had lots of trouble getting GLM-4.7-Flash to work (failed tool calls etc), and even when it works, it's at very low tok/s.”
Comment from 4 Feb 2026 in “Claude Code: connect to a local model when your quota runs out” “I recently tried GLM-4.7 Flash 30b and didn’t have a good experience with it at all.”
Comment from 31 Jan 2026 in “The $100B Megadeal Between OpenAI and Nvidia Is on Ice” “it seemed to get a bit stuck but I think I probably pushed it too far”
Comment from 20 Jan 2026 in “GLM-4.7-Flash” - Comment from 13 Jun 2026 in “GLM 5.2 Is Out”
Slow on some hardware
“I had lots of trouble getting GLM-4.7-Flash to work (failed tool calls etc), and even when it works, it's at very low tok/s.”
Comment from 4 Feb 2026 in “Claude Code: connect to a local model when your quota runs out” “glm 4.7 flash was not better, but much slower on a 16GB card”
Comment from 12 Feb 2026 in “GLM-5: From Vibe Coding to Agentic Engineering”
Other open-weight language models people recommend
The four most recommended other open-weight language models, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| Qwen 3.8 27B | 1,173 | 426 | 73% | 13 | 2 mo32 owners | Approved |
| Z.ai GLM | 1,043 | 384 | 73% | 13 | 2 mo52 owners | Approved |
| Kimi | 931 | 359 | 72% | 13 | 1 mo26 owners | Approved |
| Gemma 4 | 713 | 249 | 74% | 13 | 1 mo14 owners | Approved |
| GLM-4.7-FlashThis page | 46 | 11 | 81% | 7 | — | Approved |
What people said
40 of the 68 comments that mention it, newest first. Each links to the original comment.
“I used GLM 4.7 flash as my main model for months and it was an incredibly tenacious model and very very fast”
“I had a terrible experience with 4.7-flash.”
“one of the strongest local coding models”
“GLM 4.7 flash was my previous go to for coding”
“at least keeps making progress if I keep nudging it”
“consistently come back to GLM 4.7 because of its capabilities”
“GLM 4.7 Flash did slightly better at roughly equal speed and full run cost”
“GLM 4.7 Flash is the main model I use locally for agentic coding work, it's pretty incredible.”
“GLM 4.7 flash has been the most adept at tool calling I've found”
“a well made (as in, unsloth) smaller quant will help a good amount here, without a notable reduction in performance”
“GLM 4.7 flash is also incredibly performant on modest hardware”
“Only model that is competitive in speed is GLM 4.7 flash”
“glm 4.7 flash which is really good (to be local)”
“quite fast on the 6000, at around 100 t/s, but definitely not as smart”
“it is nowhere near as capable as its bigger sibling in use cases beyond typical chat”
“glm 4.7 flash was not better, but much slower on a 16GB card”
“And you can run quantized versions on old hardware!”
“in that space, 4.7 Flash holds its own better than it has any right to.”
“It feels something like Claude 4.5 Haiku at a parameter size where other coding models are still getting into loops”
“I had lots of trouble getting GLM-4.7-Flash to work (failed tool calls etc), and even when it works, it's at very low tok/s.”
“is a nice local model for this sort of thing”
“glm 4.7-flash that was my local model of choice”
“very few ollamas are new enough for decent modern models like glm-4.7-flash”
“I recently tried GLM-4.7 Flash 30b and didn’t have a good experience with it at all.”
“a 22gb model @q5 on m4 max pro and 59 tokens/s”
“qwen3-coder and glm 4.7 flash are both impressive 30b parameter models”
“I’m getting better results with glm-4.7-flash than with qwen3-coder:30b”
“in my experience for Claude code a 80 plus hit rate in speculative is reasonable”
“it seemed to get a bit stuck but I think I probably pushed it too far”
“it actually worked nice in opencode, which I couldn't get Nemotron to do”
“Community reports highlight strong performance in UI generation and tool calling”
“it feels dramatically worse than gpt-oss-20b”
“Comparison to GPT-OSS-20B (irrespective of how you feel that model actually performs) doesn't fill me with confidence.”
“It's really fast! But, for now it outputs garbage because there is no (good) template.”
“Seems to work okay, but there usually are subtle bugs in the implementation or chat template”
“This seems pretty darn good for a 30B model.”
“this is the perfect size for running fine-tuning experiments”
“it seems to be roughly on par with o3/GPT-5 mini, which would be pretty impressive if it translated to real-world usage”
“it seems to be roughly on par with o3/GPT-5 mini, which is pretty impressive”
“Looks like solid incremental improvements. The UI oneshot demos are a big improvement over 4.6.”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.