replace default model mistral:7b with qwen3:8b

CHANGELOG.md

@@ -32,6 +32,7 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
 - Replaced double-sort subquery in `load_history_filtered` with a CTE — eliminates redundant `ORDER BY` on the derived table (#896)
 
 ### Changed
+- Replace default Ollama model `mistral:7b` with `qwen3:8b` across config defaults, tests, snapshots, and `--init` wizard; add `"qwen3"/"qwen"` as `ChatML` aliases in `ChatTemplate::parse_str` (#897)
 - Split 3177-line `src/main.rs` into focused modules: `runner.rs` (dispatch), `agent_setup.rs` (tool/MCP/feature setup), `tracing_init.rs`, `tui_bridge.rs`, `channel.rs`, `tests.rs` — `main.rs` reduced to 26 LOC (#839)
 - Split 1791-line `crates/zeph-core/src/bootstrap.rs` into submodule directory: `config.rs`, `health.rs`, `mcp.rs`, `provider.rs`, `skills.rs`, `tests.rs` — `bootstrap/mod.rs` reduced to 278 LOC (#840)
 - Replace `source_kind: String` in `SkillTrustRow` with `SourceKind` enum (`Local`, `Hub`, `File`) with serde DB serialization; invalid values fail at parse time (#848)

config/default.toml

@@ -15,7 +15,7 @@ provider = "ollama"
 # Base URL for Ollama server
 base_url = "http://localhost:11434"
 # Primary model for chat completions
-model = "mistral:7b"
+model = "qwen3:8b"
 # Model for generating embeddings (semantic memory)
 embedding_model = "qwen3-embedding"
 
@@ -69,7 +69,7 @@ max_tokens = 4096
 # embed = "ollama"
 # [llm.orchestrator.providers.ollama]
 # type = "ollama"
-# model = "mistral:7b"          # optional: override model
+# model = "qwen3:8b"          # optional: override model
 # base_url = "http://localhost:11434"  # optional: override base URL
 # embedding_model = "qwen3-embedding"  # optional: override embedding model
 # [llm.orchestrator.providers.claude]
@@ -110,7 +110,7 @@ max_tokens = 4096
 # [llm]
 # provider = "orchestrator"
 # base_url = "http://localhost:11434"      # used by ollama sub-provider
-# model = "mistral:7b"                     # fallback model for sub-providers without explicit model
+# model = "qwen3:8b"                     # fallback model for sub-providers without explicit model
 # embedding_model = "qwen3-embedding"      # used by ollama sub-provider for embeddings
 #
 # [llm.cloud]
@@ -122,12 +122,12 @@ max_tokens = 4096
 # model = "whisper-1"
 #
 # [llm.orchestrator]
-# default = "ollama/mistral:7b"            # "provider_name/model" or just "provider_name"
+# default = "ollama/qwen3:8b"            # "provider_name/model" or just "provider_name"
 # embed = "qwen3-embedding"                # embedding model name or "provider_name/model"
 #
 # [llm.orchestrator.providers.ollama]
 # type = "ollama"
-# model = "mistral:7b"                     # optional: overrides [llm].model
+# model = "qwen3:8b"                     # optional: overrides [llm].model
 # base_url = "http://localhost:11434"      # optional: overrides [llm].base_url
 # embedding_model = "qwen3-embedding"      # optional: overrides [llm].embedding_model
 #

crates/zeph-core/src/config/snapshots/zeph_core__config__types__tests__config_default_snapshot.snap

@@ -10,7 +10,7 @@ auto_update_check = true
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 embedding_model = "qwen3-embedding"
 response_cache_enabled = false
 response_cache_ttl_secs = 3600

crates/zeph-core/src/config/tests.rs

@@ -84,7 +84,7 @@ fn defaults_when_file_missing() {
     let config = Config::default();
     assert_eq!(config.llm.provider, super::ProviderKind::Ollama);
     assert_eq!(config.llm.base_url, "http://localhost:11434");
-    assert_eq!(config.llm.model, "mistral:7b");
+    assert_eq!(config.llm.model, "qwen3:8b");
     assert_eq!(config.llm.embedding_model, "qwen3-embedding");
     assert_eq!(config.agent.name, "Zeph");
     assert_eq!(config.memory.history_limit, 50);
@@ -148,7 +148,7 @@ name = "Zeph"
 [llm]
 provider = "claude"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [llm.cloud]
 model = "claude-sonnet-4-5-20250929"
@@ -178,7 +178,7 @@ history_limit = 50
 fn env_overrides() {
     clear_env();
     let mut config = Config::default();
-    assert_eq!(config.llm.model, "mistral:7b");
+    assert_eq!(config.llm.model, "qwen3:8b");
 
     unsafe { std::env::set_var("ZEPH_LLM_MODEL", "phi3:mini") };
     config.apply_env_overrides();
@@ -202,7 +202,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -251,7 +251,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -293,7 +293,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -376,7 +376,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 embedding_model = "nomic-embed-text"
 
 [skills]
@@ -423,7 +423,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -462,7 +462,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -515,7 +515,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -567,7 +567,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -627,7 +627,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -678,7 +678,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -859,7 +859,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -907,7 +907,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -965,7 +965,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -1023,7 +1023,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -1067,7 +1067,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -1107,7 +1107,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -1432,7 +1432,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -1516,7 +1516,7 @@ name = "FullBot"
 [llm]
 provider = "claude"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 embedding_model = "nomic"
 
 [llm.cloud]
@@ -1625,7 +1625,7 @@ name = "Zeph"
 [llm]
 provider = "openai"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [llm.openai]
 base_url = "https://api.openai.com/v1"
@@ -1672,7 +1672,7 @@ name = "Zeph"
 [llm]
 provider = "openai"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [llm.openai]
 base_url = "https://api.openai.com/v1"
@@ -1723,7 +1723,7 @@ name = "Zeph"
 [llm]
 provider = "openai"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [llm.openai]
 base_url = "https://api.openai.com/v1"
@@ -1763,7 +1763,7 @@ name = "Zeph"
 [llm]
 provider = "openai"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [llm.openai]
 base_url = "https://api.openai.com/v1"
@@ -1809,7 +1809,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -1915,7 +1915,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -1959,7 +1959,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -2064,7 +2064,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -2100,7 +2100,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -2160,7 +2160,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -2196,7 +2196,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]
@@ -2241,7 +2241,7 @@ name = "Zeph"
 [llm]
 provider = "ollama"
 base_url = "http://localhost:11434"
-model = "mistral:7b"
+model = "qwen3:8b"
 
 [skills]
 paths = ["./skills"]

crates/zeph-core/src/config/types.rs

@@ -1194,7 +1194,7 @@ impl Default for Config {
             llm: LlmConfig {
                 provider: ProviderKind::Ollama,
                 base_url: "http://localhost:11434".into(),
-                model: "mistral:7b".into(),
+                model: "qwen3:8b".into(),
                 embedding_model: default_embedding_model(),
                 cloud: None,
                 openai: None,

crates/zeph-core/src/context.rs

@@ -395,14 +395,14 @@ mod tests {
             working_dir: "/tmp/test".into(),
             git_branch: Some("main".into()),
             os: "macos".into(),
-            model_name: "mistral:7b".into(),
+            model_name: "qwen3:8b".into(),
         };
         let formatted = env.format();
         assert!(formatted.starts_with("<environment>"));
         assert!(formatted.ends_with("</environment>"));
         assert!(formatted.contains("working_directory: /tmp/test"));
         assert!(formatted.contains("os: macos"));
-        assert!(formatted.contains("model: mistral:7b"));
+        assert!(formatted.contains("model: qwen3:8b"));
         assert!(formatted.contains("git_branch: main"));
     }