harn-vm 0.8.80

Async bytecode virtual machine for the Harn programming language
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99

# ---------- Together reasoning models ----------------------------------------
#
# Together exposes three reasoning surfaces today:
#   - Qwen uses `chat_template_kwargs.thinking` / `enable_thinking`.
#   - Hybrid Kimi/GLM/Gemma models use `reasoning.enabled`.
#   - GPT-OSS uses top-level `reasoning_effort`.
# Keep each host-specific rule explicit so the generic OpenAI fallback does
# not accidentally claim the wrong wire format.

[[provider.together]]
model_match = "openai/gpt-oss-*"
structured_output = "native"
thinking_modes = ["effort"]
reasoning_effort_supported = true
prefers_xml_scaffolding = false
prefers_markdown_scaffolding = true
structured_output_mode = "native_json"
supports_assistant_prefill = false
prefers_role_developer = false
prefers_xml_tools = false
thinking_block_style = "reasoning_summary"

[[provider.together]]
model_match = "deepseek-ai/deepseek-v4*"
native_tools = true
preferred_tool_format = "native"
thinking_modes = ["enabled", "effort"]
reasoning_effort_supported = true
prompt_caching = true
structured_output = "native"
server_parser = "none"
honors_chat_template_kwargs = false
prefers_xml_scaffolding = false
prefers_markdown_scaffolding = true
structured_output_mode = "native_json"
supports_assistant_prefill = false
prefers_role_developer = false
prefers_xml_tools = false
thinking_block_style = "reasoning_summary"

[[provider.together]]
model_match = "moonshotai/kimi-k2*"
native_tools = true
preferred_tool_format = "native"
structured_output = "native"
thinking_modes = ["enabled"]
honors_chat_template_kwargs = false
prefers_xml_scaffolding = false
prefers_markdown_scaffolding = true
structured_output_mode = "native_json"
supports_assistant_prefill = false
prefers_role_developer = false
prefers_xml_tools = false
thinking_block_style = "inline"

[[provider.together]]
model_match = "zai-org/glm-5*"
native_tools = true
preferred_tool_format = "native"
structured_output = "native"
thinking_modes = ["enabled"]
honors_chat_template_kwargs = false
prefers_xml_scaffolding = false
prefers_markdown_scaffolding = true
structured_output_mode = "native_json"
supports_assistant_prefill = false
prefers_role_developer = false
prefers_xml_tools = false
thinking_block_style = "inline"

[[provider.together]]
model_match = "google/gemma-4*"
native_tools = true
vision_supported = true
preferred_tool_format = "native"
structured_output = "native"
thinking_modes = ["enabled"]
honors_chat_template_kwargs = false
prefers_xml_scaffolding = false
prefers_markdown_scaffolding = true
structured_output_mode = "native_json"
supports_assistant_prefill = false
prefers_role_developer = false
prefers_xml_tools = false
thinking_block_style = "inline"

[[provider.together]]
model_match = "moonshotai/*"
native_tools = true
preferred_tool_format = "native"
structured_output = "native"
prefers_xml_scaffolding = false
prefers_markdown_scaffolding = true
structured_output_mode = "native_json"
supports_assistant_prefill = false
prefers_role_developer = false
prefers_xml_tools = false
thinking_block_style = "none"